WO2008050222A2 - Method and system for automatic quality control used in computerized analysis of ct angiography - Google Patents

Method and system for automatic quality control used in computerized analysis of ct angiography Download PDF

Info

Publication number
WO2008050222A2
WO2008050222A2 PCT/IB2007/003193 IB2007003193W WO2008050222A2 WO 2008050222 A2 WO2008050222 A2 WO 2008050222A2 IB 2007003193 W IB2007003193 W IB 2007003193W WO 2008050222 A2 WO2008050222 A2 WO 2008050222A2
Authority
WO
WIPO (PCT)
Prior art keywords
data
imaging data
aorta
identified
blood vessel
Prior art date
Application number
PCT/IB2007/003193
Other languages
French (fr)
Other versions
WO2008050222A3 (en
Inventor
Shai Levanon
Roman Goldenberg
Michael Levy
Original Assignee
Rcadia Medical Imaging Ltd.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Rcadia Medical Imaging Ltd. filed Critical Rcadia Medical Imaging Ltd.
Priority to EP07848825A priority Critical patent/EP2073701A4/en
Publication of WO2008050222A2 publication Critical patent/WO2008050222A2/en
Publication of WO2008050222A3 publication Critical patent/WO2008050222A3/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0012Biomedical image inspection
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/02Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
    • A61B5/02007Evaluating blood vessel condition, e.g. elasticity, compliance
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/48Diagnostic techniques
    • A61B6/481Diagnostic techniques involving the use of contrast agents
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/50Clinical applications
    • A61B6/504Clinical applications involving diagnosis of blood vessels, e.g. by angiography
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00ICT specially adapted for the handling or processing of medical images
    • G16H30/20ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10072Tomographic images
    • G06T2207/10081Computed x-ray tomography [CT]
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30004Biomedical image processing
    • G06T2207/30048Heart; Cardiac

Definitions

  • the field of the disclosure relates generally to computer systems. More specifically, the disclosure relates to automatic quality control used in computerized analysis of coronary CT angiography analysis.
  • a method and a system for automatic quality control (QC) for computer based analysis of medical imaging data is provided.
  • the medical imaging data is coronary computed tomography (CT) angiography (CTA) data.
  • CT coronary computed tomography
  • CTA angiography
  • the method and system provide self-diagnosis of an automatic analysis process and provide assessment of a CT angiography study quality.
  • the QC method and system automatically distinguish between cases for which a reliable result can be produced by the computerized analysis and those for which the automatic processing is impossible and/or a confidence in the result is low.
  • the QC method and system provide an estimation of the image quality, which can be used independently by medical imaging equipment operators to measure the readability of the study and to calibrate the medical imaging equipment.
  • a system for identifying a data quality issue associated with medical diagnostic data includes, but is not limited to, an imaging apparatus configured to generate imaging data and a processor operably coupled to the imaging apparatus to receive the imaging data.
  • the processor is configured to analyze the received imaging data; to identify a data quality issue associated with the received imaging data based on the analysis of the received imaging data; to generate a message including the identified data quality issue; and to send the generated message to a user.
  • a method of identifying a data quality issue associated with medical diagnostic data is provided. Imaging data is received at a computing device. The received imaging data is analyzed at the computing device. A data quality issue associated with the received imaging data is identified based on the analysis of the received imaging data at the computing device. A message including the identified data quality issue is generated at the computing device. The generated message is sent to a user.
  • computer-readable instructions are provided that, upon execution by a processor, cause the processor to implement the operations of the method of identifying a data quality issue associated with medical diagnostic data
  • FIG. 1 depicts a block diagram of an automated CT image processing system in accordance with an exemplary embodiment.
  • Figs. 2a and 2b depict a flow diagram illustrating exemplary operations performed by the automated CT image processing system of Fig. 1 in accordance with an exemplary embodiment.
  • FIG. 3 depicts a flow diagram illustrating exemplary operations performed in detecting a heart region in accordance with an exemplary embodiment.
  • FIG. 4 depicts a flow diagram illustrating exemplary operations performed in detecting a lung region in accordance with an exemplary embodiment.
  • FIGs. 5a and 5b depict a flow diagram illustrating exemplary operations performed in detecting and identifying the aorta in accordance with an exemplary embodiment.
  • Figs. 6a and 6b depict a flow diagram illustrating exemplary operations performed in identifying coronary artery vessel exit points from the aorta in accordance with an exemplary embodiment.
  • FIGs. 7a and 7b depict a flow diagram illustrating exemplary operations performed in identifying coronary artery vessel tree in accordance with an exemplary embodiment.
  • Fig. 8 depicts a flow diagram illustrating exemplary operations performed in identifying calcium seed in accordance with an exemplary embodiment.
  • Fig. 9 depicts a flow diagram illustrating exemplary operations performed in identifying in identifying soft plaque in accordance with an exemplary embodiment.
  • Fig. 10 depicts a graph illustrating X-junction removal from a coronary artery vessel tree in accordance with an exemplary embodiment.
  • FIG. 11 depicts a first user interface of a visualization application presenting summary pathology results in accordance with an exemplary embodiment.
  • Fig. 12 depicts a second user interface of the visualization application presenting multiple views of a pathology in accordance with a first exemplary embodiment.
  • Fig. 13 depicts a third user interface of the visualization application presenting a blood vessel effective lumen area as a function of distance from the aorta in accordance with an exemplary embodiment.
  • Fig. 14 depicts a fourth user interface of the visualization application presenting multiple views of a pathology in accordance with a second exemplary embodiment.
  • FIG. 15 depicts a fifth user interface of the visualization application presenting multiple views of a pathology in accordance with a third exemplary embodiment.
  • Fig. 16 depicts a flow diagram illustrating exemplary operations performed in assessing a quality of medical imaging data in accordance with an exemplary embodiment.
  • Image processing system 100 may include a computed tomography (CT) apparatus 101 and a computing device 102.
  • Computing device 102 may include a display 104, an input interface 106, a memory 108, a processor 110, a pathology identification application 112, and a visualization application 114.
  • CT apparatus 101 generates image data.
  • the present techniques are well-suited for use with a wide variety of medical diagnostic system modalities, including magnetic resonance imaging systems, ultrasound systems, positron emission tomography systems, nuclear medicine systems, etc.
  • computing device 102 may also include a communication interface, which provides an interface for receiving and transmitting data between devices using various protocols, transmission technologies, and media as known to those skilled in the art.
  • the communication interface may support communication using various transmission media that may be wired or wireless.
  • Display 104 presents information to a user of computing device 102 as known to those skilled in the art.
  • display 104 may be a thin film transistor display, a light emitting diode display, a liquid crystal display, or any of a variety of different displays.
  • Input interface 106 provides an interface for receiving information from the user for entry into computing device 102 as known to those skilled in the art.
  • Input interface 106 may use various input technologies including, but not limited to, a keyboard, a pen and touch screen, a mouse, a track ball, a touch screen, a keypad, one or more buttons, etc. to allow the user to enter information into computing device 102 or to make selections presented in a user interface displayed on display 104.
  • Input interface 106 may provide both an input and an output interface. For example, a touch screen both allows user input and presents output to the user.
  • Memory 108 is an electronic holding place or storage for information so that the information can be accessed by processor 110 as known to those skilled in the art.
  • Computing device 102 may have one or more memories that use the same or a different memory technology. Memory technologies include, but are not limited to, any type of RAM, any type of ROM, any type of flash memory, etc.
  • Computing device 102 also may have one or more drives that support the loading of a memory media such as a compact disk or digital video disk.
  • Processor 110 executes instructions as known to those skilled in the art.
  • the instructions may be carried out by a special purpose computer, logic circuits, or hardware circuits.
  • processor 110 may be implemented in hardware, firmware, software, or any combination of these methods.
  • execution is the process of running an application or the carrying out of the operation called for by an instruction.
  • the instructions may be written using one or more programming language, scripting language, assembly language, etc.
  • Processor 110 executes an instruction, meaning that it performs the operations called for by that instruction.
  • Processor 110 operably couples with display 104, with input interface 106, with memory 108, and with the communication interface to receive, to send, and to process information.
  • Processor 110 may retrieve a set of instructions from a permanent memory device and copy the instructions in an executable form to a temporary memory device that is generally some form of RAM.
  • Computing device 102 may include a plurality of processors that use the same or a different processing technology.
  • Pathology identification application 112 performs operations associated with the analysis imaging data to identify and evaluate blood vessel structures and to identify and evaluate pathologies associated with the identified blood vessel structures. Some or all of the operations and interfaces subsequently described may be embodied in pathology identification application 112. The operations may be implemented using hardware, firmware, software, or any combination of these methods. With reference to the exemplary embodiment of Fig. 1 , pathology identification application 112 is implemented in software stored in memory 108 and accessible by processor 110 for execution of the instructions that embody the operations of pathology identification application 112. Pathology identification application 112 may be written using one or more programming languages, assembly languages, scripting languages, etc. Pathology identification application 112 may integrate with or otherwise interact with visualization application 114.
  • Visualization application 114 performs operations associated with presentation of the blood vessel analysis and identification results to a user. Some or all of the operations and interfaces subsequently described may be embodied in visualization application 114. The operations may be implemented using hardware, firmware, software, or any combination of these methods. With reference to the exemplary embodiment of Fig. 1 , visualization application 114 is implemented in software stored in memory 108 and accessible by processor 110 for execution of the instructions that embody the operations of visualization application 114. Visualization application 114 may be written using one or more programming languages, assembly languages, scripting languages, etc.
  • CT apparatus 101 and computing device 102 may be integrated into a single system such as a CT imaging machine.
  • CT apparatus 101 and computing device 102 may be connected directly.
  • CT apparatus 101 may connect to computing device 102 using a cable for transmitting information between CT apparatus 101 and computing device 102.
  • CT apparatus 101 may connect to computing device 102 using a network.
  • computing device 102 is connected to a hospital computer network and a picture archive, and communication system (PACS) and receives a CT study acquired on CT apparatus 101.
  • PACS picture archive, and communication system
  • CT images are stored electronically and accessed using computing device 102.
  • CT apparatus 101 and computing device 102 may not be connected. Instead, the CT study acquired on CT apparatus 101 may be manually provided to computing device 102.
  • CT study may be stored on electronic media such as a compact disc or a digital video disc.
  • computing device 102 may initiate automatic processing of the set of images that comprise the CT study.
  • CT apparatus 101 is a 64-slice multi-detector advanced CT scanner having a reconstructed slice width and inter-slice distance less than or equal to approximately 0.5 millimeters (mm), which produces standard digital imaging and communications in medicine (DICOM) images.
  • Computing device 102 may be a computer of any form factor.
  • Image processing system 100 may provide an initial classification and decision support system, which allows fast and accurate ruling out of the three major diseases associated with chest pain.
  • Image processing system 100 can be provided as a primary CT study inspection tool assisting an emergency room physician. Additionally, image processing system 100 can be used either to completely rule out some or all of the three diseases (in case of negative results) or as a trigger to call a radiologist and/or a cardiologist to further analyze the case. Additionally, image processing system 100 may automatically identify and segment blood vessel trees and automatically analyze each blood vessel to detect and map all relevant pathologies, including calcified and soft plaque lesions and their degree of stenosis.
  • pathology identification application 112 receives CT image data.
  • the CT image data may be received from CT apparatus 101 directly or using a network.
  • the CT image data also may be received using a memory medium.
  • a heart region is identified from the received CT image data.
  • Data associated with the identified heart region is stored at computing device 102.
  • the data associated with the identified heart region includes a heart bounding box.
  • exemplary operations associated with identifying the heart region data are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. For larger studies that may include head and neck or abdominal regions, a determination of the heart region provides a correct anatomical starting point for further segmentation. For smaller studies, a determination of the heart region reduces the size of the processed region and removes the non-relevant areas to reduce the false alarm risk. In an exemplary embodiment, the top and bottom boundaries of the heart region are cut a predetermined distance above and below the heart center.
  • a first projection into an X-Y plane is defined.
  • a positive X-axis is defined as extending out from the left side of the body.
  • a positive Y-axis is defined as extending out from the back side of the body.
  • a positive Z-axis is defined as extending out from the head of the body.
  • the first projection is defined by summing the DICOM series along the Z-axis.
  • a threshold is applied to the first projection. For example, a threshold greater than approximately zero may be applied to eliminate the negative values of air which dominate the region outside the heart region and to retain the positive Hounsfeld unit (HU) values which include the fat, blood, and bones within the heart region.
  • a first largest connect component (CC) is identified in the thresholded first projection.
  • a first center of mass of the first largest CC is determined and denoted as X c , Y c i.
  • a second projection into a Y-Z plane is defined.
  • the second projection is defined by summing the DICOM series along the X-axis.
  • a threshold is applied to the second projection data. For example, a threshold greater than approximately zero may be applied to eliminate the negative values of air which dominate the region outside the heart region and to retain the positive HU values which include the fat, blood, and bones within the heart region.
  • a second largest CC is identified in the thresholded second projection data.
  • a second center of mass of the second largest CC is determined and denoted as Y C2 , Z 0 .
  • a heart region center is defined as X 0 , Yd, Z 0 .
  • a heart region bounding box is defined from the heart region center and an average heart region width in each axis direction, Wx, W ⁇ , W z .
  • the defined heart region is compared to a quality control parameter associated with a valid heart region. For example, the size and the position of the defined heart region is compared with an expected size and position within the CT study. Additionally, a tissue type, a contrast, and arterials inside the defined heart region may be compared with the expected tissue type, contrast, and arterials inside a heart region.
  • a determination is made concerning whether or not the heart region data is valid. If the heart region data is valid, processing continues at an operation 322. If the heart region data is not valid, processing continues at an operation 324. In operation 322, the defined heart region bounding box is stored at computing device 102.
  • a message is output to the user.
  • a message may be printed from computing device 102, may be displayed at computing device 102, may played audibly from computing device 102, may be transmitted to the user (i.e., e-mail, text message, instant message, voice message, etc.), etc.
  • the message may include an error indication that identifies a source of the invalidity or unreliability determination.
  • the message may indicate an invalidity of the heart region determination.
  • processing of the CT image data stops.
  • a lung region is identified from the received CT image data.
  • Data associated with the identified lung region is stored at computing device 102.
  • the data associated with the identified lung region may include a lung mask.
  • exemplary operations associated with identifying the lung region data are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting.
  • a lung threshold is applied to each slice of the DICOM series data. For example, a lung threshold of -400 HU may be applied.
  • a morphological filter is applied to the binary image.
  • the binary image is filtered using a morphological closing operation to define a lung region in the CT image data. Other processes for filling holes in the image may be used as known to those skilled in the art.
  • a lung region is defined.
  • the defined lung region is compared to a quality control parameter associated with a valid lung region. For example, the intensity levels and patterns inside the defined lung region are compared with expected intensity levels and patterns in an lung region.
  • a thorax region is identified from the received CT image data. Data associated with the identified thorax region is stored at computing device 102.
  • the thorax region may be defined as a convex hull of the lungs and the diaphragm.
  • the pulmonary arteries are identified from the received CT image data. Data associated with the identified pulmonary arteries are stored at computing device 102.
  • the received CT image data is preprocessed.
  • preprocessing may include image enhancement, smoothing, noise reduction, acquisition artifacts detection, etc.
  • image enhancement algorithms include Gaussian smoothing, median filtering, bilateral filtering, anisotropic diffusion, etc.
  • image quality test data is calculated to determine a quality of the image.
  • the calculated image quality test data is compared to quality control parameters associated with quality image data.
  • a determination is made concerning whether or not the image quality is satisfactory.
  • An exemplary process for evaluating a quality of the image data is shown with reference to Fig. 16. If the image quality is satisfactory, processing continues at an operation 214. If the image quality is unsatisfactory, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an unsatisfactory image quality.
  • the left and right main pulmonary artery trees are defined.
  • the lumen of the left and right main pulmonary artery trees is analyzed to identify any pulmonary embolism candidates.
  • any pulmonary embolism candidates are classified.
  • possible pulmonary embolism lesions are identified.
  • the ascending and the visible part of the abdominal aorta are segmented.
  • the lumen of the abdominal aorta is segmented.
  • a 3-D geometry of the abdominal aorta is modeled.
  • the modeled aorta is compared to a hypothesized "normal" abdominal aorta to identify deviations from the hypothesized "normal” abdominal aorta.
  • suspicious locations are detected and analyzed to identify dissections and aneurysms.
  • the aorta is identified.
  • the aorta is detected in the first imaging slice of the heart region bounding box based, for example, on intensity and shape properties including circularity, compactness, and area.
  • the remainder of the aorta is identified by moving from slice to slice and looking for a similar 2-D object in each slice.
  • Data associated with the identified aorta is stored at computing device 102.
  • the data associated with the identified aorta includes an aorta shape and boundary in the identified heart region. It is assumed that the heart region bounding box detected on the previous step includes the aorta exit from the heart and that the cross section of the aorta in the upper slice of the heart region is approximately circular.
  • a first slice is selected from the heart region data.
  • an aorta threshold is applied to the first slice of the DICOM series data. For example, a lung threshold of 200 HU may be used.
  • a morphological filter is applied to the binary image.
  • the binary image is filtered using a series of morphological filtering operators including an opening operator using a first parameter.
  • one or more CCs are identified.
  • a compactness of each identified CC is determined.
  • identified CCs having a determined compactness that exceeds a compactness threshold are eliminated from further consideration.
  • the compactness threshold is 0.75.
  • a size of each identified connected component is determined.
  • identified CCs having a size that exceeds a maximum size threshold or that is below a minimum size threshold are eliminated from further consideration.
  • the maximum size threshold is 10,000 pixels.
  • the minimum size threshold is 1 ,000 pixels.
  • a determination is made concerning whether or not any identified CCs remain for consideration. If identified CCs remain for consideration, processing continues at an operation 518. If no identified CCs remain for consideration, processing continues at an operation 520. In operation 518, an initial aorta candidate is selected from the remaining CCs. For example, if a plurality of identified CCs remain for consideration, the largest candidate CC that is foremost in the body is selected as the initial aorta candidate.
  • a next slice is selected from the heart region data.
  • the aorta threshold is applied to the next slice of the DICOM series data.
  • the morphological filter is applied to the binary image.
  • one or more CCs are identified.
  • a compactness of each identified CC is determined.
  • the identified CCs having a determined compactness that exceeds the compactness threshold are eliminated from further consideration.
  • a size of each identified connected component is determined.
  • the identified CCs having a size that exceeds the maximum size threshold or that is below the minimum size threshold are eliminated from further consideration.
  • a determination is made concerning whether or not any identified CCs remain for consideration in the current slice. If identified CCs remain for consideration, processing continues at an operation 538. If no identified CCs remain for consideration, processing continues at operation 520.
  • the identified CCs from the current slice are compared with the aorta candidate object(s) created from the previous slice(s).
  • a determination is made concerning whether or not any identified CCs match CCs identified from the previous slices.
  • the matched CC is assigned to the aorta candidate object. For example, if a center of a CC is closer than twenty pixels to the center of an aorta candidate object, the CC may be identified as matched with the aorta candidate object.
  • a new aorta candidate object is created based on the CC.
  • an operation 546 a determination is made concerning whether or not all of the slices have been processed. If slices remain, processing continues at operation 520. If no slices remain, in an operation 548, aorta candidate objects are eliminated based on length. For example, aorta candidate objects that persist for less than 20 slices may be removed from further consideration.
  • an aorta object is selected from the remaining aorta candidate objects. For example, the aorta candidate object closest to the upper left corner of the image may be selected as the aorta object.
  • a bounding box is defined around the selected aorta object to identify a region in which the aorta is located in the CT image data.
  • test data associated with the selected aorta object are calculated. For example, geometrical parameters associated with the selected aorta object and its relative position inside the heart bounding box are measured and/or calculated as test data.
  • the calculated test data are compared to a quality control parameter associated with a valid aorta object. For example, the diameter of the aorta should not change too much from the aortic arch to the entry point to the heart. Additionally, aorta cross sections should be approximately circular with a diameter within a valid range such as from 20 mm to 50 mm. Some or all of these parameters may be used as quality control parameters associated with determining if a selected aorta object is valid.
  • exit points of the coronary arteries from the aorta are identified by evaluating all structures connected to the aorta object which look like a vessel.
  • the direction of the vessel near a link point with the aorta object should be roughly perpendicular to an aorta centerline.
  • the left and right coronary arteries are expected to exit from the aorta object in a certain direction relative to the body coordinate system. If there are several exit point candidates, the exit point candidate which leads to a larger blood vessel tree is selected. It is assumed that the selected aorta object includes the points where the left and right coronary trees connect with the aorta.
  • Imaging slices including the aorta bounding box and a mask of the aorta detected at a previous slice are processed to detect the aorta at the current slice.
  • a first slice is selected from the aorta object.
  • regions are segmented based on a segmentation threshold at the detected aorta edges.
  • the segmentation threshold may be calculated from the median value of pixels of the smoothed image at the detected edges.
  • a ring of pre-defined radius is defined around the aorta edges detected on the previous slice and the edges are found in the ring on the current slice. Small edges are removed from further consideration.
  • the segmentation threshold may be selected adaptively. For example, if no edges are found in the ring using the calculated segmentation threshold, the calculated segmentation threshold is reduced by half, and the procedure is repeated. If no edges are found using the lowered segmentation threshold, the calculated segmentation threshold is used for subsequent slices.
  • the segmented image is post-processed. For example, small segmented objects are removed, possible vessels are removed from the segmented aorta candidates, and the segmented aorta candidates are intersected with the aorta detected in the previous slice.
  • the aorta candidates are validated by ensuring that there is at least one candidate that intersected the aorta detected in the previous slice and by ensuring that the aorta does not grow too fast. For example, if the aorta size in both a previous and a current slice is larger than 1500 pixels, the size growth ratio may be limited to 1.4.
  • the aorta candidates are selected.
  • CCs with a small intersection with the previously detected aorta are removed from consideration, and the upper-left-most candidate is chosen if a plurality of aorta candidates exist in the current slice.
  • the compactness of the selected aorta candidate is checked to ensure that the candidate is not compact. If the aorta candidate is not compact, the aorta search window is limited for the next slice. If the aorta candidate is compact, the whole image is used to search for the aorta in the next slice.
  • a bounding box for the aorta is calculated.
  • the bounding box size may be fixed and only the position of the bounding box updated to compensate for aorta movement. If the aorta candidate is compact, the bounding box may be attached to the upper left side of the aorta.
  • a vesselness score is calculated for each voxel of the aorta object.
  • the vesselness score can be determined using a vesselness function.
  • a vesselness function is a widely used function based on the analysis of Hessian eigen values.
  • a good description of an exemplary vesselness function can be found for example in Frangi, A. F., Niessen, W. J., Vincken, K. L. and Viergever, M. A., 1998, "Multiscale Vessel Enhancement Filtering", MICCAI'98, LNCS 1496, pp. 130-137.
  • a vesselness threshold is applied to the calculated vesselness score to identify possible exit points based on the HU value of a ring around the aorta object. Pixels in a ring around the detected aorta are grouped into CCs, which are analyzed to choose the most probable candidates for coronary tree exit points.
  • possible exit points are filtered to remove false candidates. For example, the possible exit points may be filtered based on a size of the CC corresponding to the possible exit, a location of the CC relative to the aorta, an incident angle of the CC relative to the aorta, etc.
  • a determination is made concerning whether or not any exit points are left.
  • a CC is identified for each exit point included in the possible exit points list.
  • a volume is calculated for each exit point CC (EPCC).
  • EPCC exit point CC
  • any EPCC having a volume below a volume threshold is eliminated from further consideration as an exit point.
  • the volume threshold may be 1500 voxels.
  • a maximum width of each EPCC is calculated.
  • any EPCC having a maximum width below a width threshold is eliminated from further consideration as an exit point.
  • the width threshold may be 2 mm.
  • a distance to the aorta is calculated for each EPCC.
  • the EPCC having a minimum distance to the aorta is selected.
  • a determination is made concerning whether or not a plurality of EPCCs remain. If a plurality of EPCCs remain, processing continues at an operation 644. If a plurality of EPCCs do not remain, processing continues at an operation 646.
  • the EPCC having a maximum width is selected from the plurality of EPCCs remaining.
  • the exit point is identified from the selected EPCCs.
  • the geometry of the selected EPCC including a change of the vessel radius and the position and orientation of the selected EPCC in the heart bounding box are analyzed.
  • the analysis data are compared to a quality control parameter associated with a valid EPCC object.
  • a valid EPCC object should extend for a certain minimal distance from the exit point. An exemplary minimum distance is two centimeters.
  • the EPCC object is indeed a tree of vessels, its convex hull to volume ratio also should be relatively low.
  • the width of the structures in the EPCC object also should be low compared to a maximal expected radius of the blood vessel such as three mm.
  • the EPCC corresponding to the right coronary tree should approximately cover the statistically expected right coronary tree region. The same should be true for the left coronary tree. Some or all of these parameters may be used as quality control parameters associated with determining if a selected EPCC object is valid. In an operation 652, a determination is made concerning whether or not the selected EPCC object is valid. If the selected EPCC object is valid, processing continues at an operation 654. If the selected EPCC object is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the EPCC determination.
  • a coronary artery vessel tree is defined. For a traversed section of the blood vessel tree, a set of end points is identified, and an attempt is made to continue tracking beyond the end point in the direction of the corresponding tree branch. If an additional tree segment is detected, it is connected to the traversed tree, and the processing continues recursively. The process is finished when no branch can be continued.
  • the stopping condition may result in connecting a wrong structure to the coronary tree (e.g. a vein or some debris in a noisy CT study).
  • the successfully tracked vessels are identified and those which remain to be detected are identified. For example, which blood vessels are to be segmented (e.g.
  • the left coronary artery (LCA), the right coronary artery (RCA), the left main coronary artery (LM), the middle portion of the left anterior descending coronary artery (LAD), and the proximal portion of the left circumflex coronary artery (LCX) and others) may be defined as an input to the process.
  • a maximum vessel length to track (e.g. five centimeters from the aorta) and a minimum blood vessel diameter to continue tracking also may be defined as inputs to the process.
  • the location of some blood vessels may be based on anatomical landmarks.
  • the RCA is located in the right atrioventricular plane.
  • a graph representation of the segmented vessel tree can be built from the identified end points and bifurcation points.
  • Graph nodes are the end points and the bifurcation points.
  • the edges are segments of a vessel centerline between the nodes.
  • exemplary operations associated with defining the coronary artery vessel tree are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting.
  • the vesselness score data is received.
  • a first binary volume is defined for a first threshold.
  • the first binary volume includes a '1' for each voxel that exceeds the first threshold and a '0' for each voxel that does not exceed the first threshold.
  • a second binary volume is defined for a second threshold.
  • the second binary volume includes a '1' for each voxel that exceeds the second threshold and a '0' for each voxel that does not exceed the second threshold.
  • the second threshold has a higher HU value than the first threshold.
  • one or more CCs in the first binary volume that intersect voxels from the second binary volume are selected as the one or more vessel CCs (VCCs). In an exemplary embodiment, intersection may be determined based on a spatial proximity between the CCs.
  • the first threshold and the second threshold are selected based on a statistical analysis of the input data such that the amount of voxels above the first threshold is approximately 0.15% of the total number of voxels in the volume and such that the amount of voxels above the second threshold is approximately 0.45% of the total number of voxels.
  • the selected VCCs are labeled in the first binary volume.
  • a starting point or root is selected for a first VCC.
  • a width map is calculated for the first VCC.
  • the width map includes the width of the VCC or the inverse distance from any point in the VCC to the boundary of the VCC. Thus, small values are near the centerline of the VCC and larger values are at the edges of the VCC.
  • the exit point of the selected VCC is identified in the binary volume.
  • the RCA tree has a single exit point. Additionally, the LM tree has a single exit point.
  • the exit point of the selected VCC is compared to a quality control parameter associated with a valid exit point. For example, a relative position along the aorta of the exit point of the selected VCC and a size of the tubular structure components connected to the selected VCC are compared with expected values. Additionally, statistics on the positions of certain blood vessels relative to the heart related coordinate system may be collected over a large database of studies. Any given VCC may be compared to the derived statistically expected location of the blood vessels and the likelihood of the VCC being that vessel is computed. Some or all of these parameters may be used as quality control parameters associated with determining if a selected exit point is valid. In an operation 716, a determination is made concerning whether or not the exit point is valid. If the exit point is valid, processing continues at an operation 717. If the exit point is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the exit point determination.
  • a distance map is calculated for the first VCC.
  • the distance is calculated from any point in the VCC to the identified exit point.
  • the distance map includes the calculated distance weighted by the width to ensure that the minimal path follows the vessel centerline.
  • candidate endpoints are identified. For example, during the calculation of the weighted distance map, one or more candidate endpoints may be saved. The candidate endpoints are voxels, which did not update any of their neighbors during the distance map calculation.
  • non-local maxima candidate endpoints are filtered. Thus, the candidate endpoints are scanned and only local maxima with respect to the distance from the root over a given window are kept. This process eliminates candidates that are not true blob vessel end points.
  • an identifier for the candidate endpoint is created.
  • a new vertex is added to a symbolic graph of the vessel tree.
  • a first candidate endpoint is backtracked to the root to create graph edges.
  • An auxiliary volume is used to mark voxels that have already been visited.
  • the back-tracking may be a gradient descent iterative process (the gradient is in the distance field). Because the weighted distance map contains a single global minimum (the root), convergence is guaranteed. The method used to define the distance map ensures that the backtracking will be along the centerline or close to it.
  • all visited voxels in the auxiliary volume are marked with the current vertex identifier during the backtracking.
  • an operation 730 a determination is made concerning whether or not a root is reached. If a root is reached, processing continues at an operation 732. If a root is not reached, processing continues at an operation 734. In operation 732, a new edge and vessel path are defined based on the backtracking. Processing continues at an operation 740. In an operation 734, a determination is made concerning whether or not an already visited voxel is reached. If an already visited voxel is reached, processing continues at an operation 736. If an already visited voxel is not reached, processing continues at an operation 726 to continue the backtracking to the endpoint. In an operation 736, a new edge and a new bifurcation vertex are defined. In an operation 738, the new bifurcation vertex is connected to the currently backtracked path, and the new edge and the new bifurcation vertex are added to the vessel tree.
  • processing continues at an operation 746.
  • short branches are removed based on the rationale that they do not contribute to the analysis because important findings are usually located at the major blood vessels, which are thick and elongated. Therefore, in an exemplary embodiment, graph edges which lead to endpoints that are less than a length threshold are removed.
  • An exemplary length threshold is five mm.
  • x-junctions are removed to eliminate veins. Veins are usually faint and spatially close to the arteries. X-junctions are defined as two very close bifurcation points. For example, close bifurcation points may be less than approximately 3 mm from each other.
  • a bifurcation may be two VCCs intersecting at angles between approximately 70 degrees and approximately 110 degrees. Additionally, a bifurcation may be two VCCs intersecting at angles approximately equal to 90 degrees. The sub-tree which remains is the one which has the closest direction to the edge arriving from the aorta.
  • a vessel tree 1000 includes a first vessel path 1002, a second vessel path 1004, and a third vessel path 1006.
  • First vessel path 1002 and second vessel path 1004 form a first x-junction 1008.
  • First vessel path 1002 has the closest direction to the edge arriving from the aorta and is selected to remain in the vessel tree.
  • Second vessel path 1004 is removed.
  • First vessel path 1002 and third vessel path 1004 form a second x-junction 1010.
  • first vessel path 1002 has the closest direction to the edge arriving from the aorta and is selected to remain in the vessel tree.
  • Third vessel path 1006 is removed.
  • an operation 750 single entry, single exit point vertices are removed. These vertices are created when one of the endpoints is recursively continued. The vertex is removed, and the two edges are joined to a single path.
  • a determination is made concerning whether or not all VCCs have been processed. If all VCCs have been processed, processing continues at an operation 754. If all VCCs have not been processed, processing continues at operation 710.
  • blood vessel test data are calculated. For example, a volume and a length of the detected blood vessels is calculated.
  • the calculated blood vessel test data are compared to a quality control parameter associated with a valid blood vessel tree.
  • the presence of blood vessels in the expected anatomical locations such as the atrioventricular grooves and inter-ventricular plane is evaluated.
  • the RCA is expected to go through the atrioventricular groove between the right ventricle and the right atrium.
  • the groove may be automatically detected, and the RCA candidate vessel's location compared to the groove mask. If a significant part of the vessel is located outside the groove, or, alternatively, if a large part of the groove is not covered by the vessel, the vessel may be determined not to be the RCA.
  • a determination is made concerning whether or not the blood vessel tree is valid. If the blood vessel tree is valid, processing continues at an operation 760. In operation 760, processing continues. If the exit point is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the blood vessel tree determination.
  • the defined coronary artery vessel tree is labeled.
  • a list of graph edges may be assigned to each blood vessel tracked by analyzing the relative section positions and locations relative to detected anatomical heart landmarks.
  • blood vessel test data are analyzed to determine a validity of the coronary artery tree labeling process. For example, the presence of the major coronary arteries such as LCA, RCA, LM, LAD, and LCX and their relative positions are analyzed.
  • the analyzed data is compared to a quality control parameter associated with valid blood vessel positions.
  • RCA and the LCX are approximately in one plane (RCA-LCX plane) and that the remaining blood vessels are on one side of the RCA-LCX plane. It further is expected that the LAD is approximately perpendicular to the RCA-LCX plane.
  • a determination is made concerning whether or not the blood vessel labeling process is valid. If the blood vessel labeling process is valid, processing continues at an operation 246. If the blood vessel labeling process is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the blood vessel labeling process.
  • the blood vessel tree represented by a centerline for each blood vessel segment is stored at computing device 102.
  • a radius of each blood vessel is determined.
  • a blood vessel centerline is determined.
  • "Sausages" of blood vessels are obtained from the coronary artery vessel tree. Each "sausage” includes axial blood vessel cross-sections taken perpendicular to the blood vessel direction. Initially, a blood vessel center is presumed to be at the center of each section. Either “stretched” or “curved” blood vessels can be used. Any blood vessel radius and center line estimation method can be used. In an exemplary embodiment, low pass post-filtering between consecutive cross-sections is performed. Because a blood vessel may be surrounded by tissue having similar attenuation values, indirect indicators may be used to define the blood vessel edge.
  • a blood vessel center consisting of a number of pixels can be defined, in particular when a cross section is elongated. In an exemplary embodiment, the blood vessel center is reduced to a single pixel.
  • an operation 250 areas of calcium are identified in each blood vessel. Any high precision calcium identification method can be used.
  • a cross section based analysis aimed at location of the calcium seeds is performed, and a volume based analysis aimed at removal of spurious seeds created by the "salt noise" and by the growing of valid seeds into the correct calcium area is performed.
  • exemplary operations associated with identifying the areas of calcium, if any, in each blood vessel are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting.
  • a first slice is selected from the heart region data.
  • a calcium threshold is applied to the first slice.
  • the calcium threshold is 150 intensity levels above the blood vessel lumen level. Adaptive threshold values taking into account expected lumen values are used.
  • a morphological filter is applied to the thresholded first slice based on the non-concentric nature of the calcium deposits. Empirical observations indicate that calcium tends to appear close to the blood vessel borders.
  • a maximum intensity in a given cross-section is identified as a possible location of a calcium seed.
  • a distance from the center to the maximum intensity is calculated.
  • a determination is made concerning whether or not the calculated distance exceeds a calcium distance threshold.
  • the calcium distance threshold is based on a comparison with an estimated radius value. If the distance does not exceed the calcium distance threshold, processing continues in an operation 814. If the distance does exceed the calcium distance threshold, processing continues in an operation 812. In operation 812, an area of the calcium seed is calculated. In operation 814, a determination is made concerning whether or not any vessels remain for processing. If vessels remain, processing continues at operation 808. If no vessels remain, processing continues at an operation 816. In operation 816, a determination concerning whether or not any slices remain for processing is performed. If no slices remain, processing continues at an operation 820. If slices remain, processing continues at an operation 818. In operation 818, the next slice is selected from the heart region data and processing continues at operation 802.
  • a volume of any identified calcium seed(s) is calculated based on the area calculated for each slice and the number of slices over which the identified calcium seed(s) extends. If a calcium seed is identified, it also is extended to the surrounding high intensity areas providing that no "spill to the center" occurs. An extent of the calcium seed may be determined based on a threshold. For example, lumen intensities exceeding approximately 650 HU may be considered to be calcified plaque or part of the calcium seed.
  • an operation 252 areas of soft plaque are identified in each blood vessel by the low intensity inside the blood vessel area. Any high precision soft plaque identification method can be used. With reference to Fig. 9, exemplary operations associated with identifying the areas of soft plaque, if any, in each blood vessel are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting.
  • a first slice is selected from the heart region data.
  • a soft plaque threshold is applied to the first slice. In an exemplary embodiment, the soft plaque threshold is between approximately 50 HU and approximately 200 HU.
  • Adaptive threshold values taking into account expected lumen values are used in an exemplary embodiment.
  • a determination is made concerning whether or not calcium is present.
  • the presence of calcium may indicate the presence of the frequent figure "8" shaped pattern.
  • calcium is located in one of the ovals of the "8".
  • a lumen is located in the other oval of the "8", and soft plaque connects the two ovals. If calcium is present, processing continues at an operation 906. If calcium is not present, processing continues at an operation 908. In operation 906, a soft plaque area is identified and processing continues at an operation 911.
  • a volume of any identified soft plaque area(s) is calculated based on the area calculated for each slice and the number of slices over which the identified soft plaque area(s) extends.
  • a volume of any identified calcium seed(s) is updated to include areas between the calcium seed and the blood vessel border and between the calcium and soft plaque areas to compensate for natural intensity low passing that may have occurred during the CT image acquisition.
  • a severity of any obstructions identified containing soft plaque or calcium is calculated. Any obstruction computation method can be used.
  • a total obstruction ratio is calculated as a ratio of the total calcium and soft plaque areas divided by the total blood vessel area excluding the border area.
  • an obstruction is identified to be severe if the total obstruction ratio exceeds 50% for at least two consecutive cross sections. An examining physician may be allowed to control the threshold to achieve a system sensitivity matching their clinical requirements.
  • the cross section images may appear reasonably normal.
  • pathology must be identified based on the analysis of global variations.
  • global filters are applied to identify pathologies. For example, a first filter may be applied to identify a rapid decrease in the blood vessel radius. A second filter may be applied to identify a rapid decrease in the lumen intensity. A third filter may be applied to identify a rapid increase in the lumen intensity.
  • the decisions from the series of filters may be cumulative. As a result, it is sufficient if a pathology is identified through use of one of the three filters.
  • the filters may use the values of blood vessel radius and luminance as computed above.
  • pathology test data are calculated to determine a validity of the pathology identification process. For example, the number of identified pathologies per blood vessel segment is calculated.
  • the calculated data is compared to a quality control parameter associated with valid pathology test data. For example, a statistically expected number of detected pathologies per blood vessel segment for a specific groups of patients is determined and compared with the calculated pathology test data.
  • a determination is made concerning whether or not the pathology identification process is valid. For example, a significant deviation from the statistically expected number of detected pathologies per blood vessel segment for a specific groups of patients indicates invalidity. If the pathology identification process is valid, processing continues at an operation 264. If the pathology identification process is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the pathology identification process.
  • First user interface 1100 may include a header portion 1102.
  • Header portion 1102 may include patient data, study data, and/or acquisition data.
  • data displayed in header portion 1102 may be obtained from a header of the DICOM data.
  • First user interface 1100 further may include a blood vessel list portion 1104.
  • Blood vessel list portion 1104 may include a list of the blood vessels in the created blood vessel tree.
  • Displayed next to a name identifying each blood vessel may be information related to each blood vessel including a lumen status, a total number of lesions, a number of calcium lesions, and a number of soft plaque lesions.
  • the lumen status may indicate "normal” or a percentage of blockage that may be a percentage range. If a plurality of lesions are present, the range may indicate the maximum blockage range.
  • a maximum volume and Agatston score may be displayed for the calcium lesions.
  • a maximum volume and score also may be displayed for the soft plaque lesions.
  • User selection of a blood vessel 1106 in blood vessel list portion 1104 may cause display of a detailed description of the lesions associated with the selected blood vessel in a detail portion 1108.
  • Detail portion 1108 may include a list of the lesions. For each lesion, a segment name, a lesion type, a degree of stenosis value, a volume, a distance from the aorta, a distance from the blood vessel origin, an eccentricity, a degree of positive remodeling, and a morphological regularity may be shown.
  • First user interface 1100 further may include a totals portion 1110.
  • Totals portion 1110 may include summary data associated with a degree of stenosis, lesions, the number of stents, etc.
  • a visualization of the blood vessels is provided to a user based on the processes described with reference to Figs. 2-9.
  • a second user interface 1200 of visualization application 114 in accordance with a first exemplary embodiment is shown.
  • four simultaneous views of the same pathology may be shown to facilitate a correct diagnosis with each view presented in a different area of second user interface 1200.
  • Each view may be created using a variety of graphical user interface techniques in a common window, in separate windows, or in any combination of windows.
  • Second user interface 1200 may include a first axial slice viewer 1202, a first 3-D coronary vessel map 1204, a first stretched blood vessel image 1206, and an intra-vascular ultrasound (IVUS) type view 1208.
  • First axial slice viewer 1202 presents intensity levels from a slice of imaging data. The intensity levels may be indicated in color or gray-scale. For example, first axial slice viewer 1202 may indicate an identified pathology 1203 in red. First axial slice viewer 1202 may present the slice of imaging data in a top left area of second user interface 1200.
  • First 3-D coronary vessel map 1204 provides a view of the blood vessel tree synchronized with first axial slice viewer 1202 to indicate the identified pathology 1203.
  • First 3-D coronary vessel map 1204 may be presented in a top right area of second user interface 1200 and may include a 3-D grid to identify the length of the blood vessels in the blood vessel tree in each direction. Selecting an area of first stretched blood vessel image 1206 may cause image rotation of first 3-D coronary vessel map 1204 around its axis to facilitate a correct 3-D perception of the blood vessel structure.
  • First 3-D coronary vessel map 1204 may be synchronized with first axial slice viewer 1202 to distinguish the selected blood vessel from the remaining blood vessels in the blood vessel tree.
  • First 3-D coronary vessel map 1204 may be rotated using an input interface as known to those skilled in the art.
  • Indicators may be provided in first 3-D coronary vessel map 1204 to indicate end points and bifurcations. For example, end points may be indicated using green circles and bifurcations may be indicated using red circles.
  • First stretched blood vessel image 1206 includes a vertical bar which denotes a location of the slice displayed in first axial slice viewer 1202 in a stretched view of a selected blood vessel.
  • First stretched blood vessel image 1206 may be located in a bottom left area of second user interface 1200.
  • the physician can superimpose corresponding plaque areas. For example, soft plaque may be indicated in red and calcified plaque indicated in blue. If desired, the physician can invoke an edit mode and correct automatic results.
  • a third user interface of visualization application 114 graphically displays a lumen area of each blood vessel to clearly identify all stenosis lesions and to allow an evaluation of their severity.
  • the graphical display includes a distance from the aorta on the X-axis and a lumen area on the Y-axis.
  • a first curve 1300 indicates a normal blood vessel lumen.
  • a second curve 1302 indicates a stenosis due to calcified plaque.
  • a third curve 1304 indicates a stenosis due to soft plaque.
  • Fourth user interface 1400 may include a second axial slice viewer 1402, a second 3-D coronary vessel map 1404, a second stretched blood vessel image 1406, and a pathology report type view 1408.
  • Second axial slice viewer 1402 may include an axial slice of the imaging data presented in a top left area of second user interface 1400. provides a current location on the 3-D coronary vessel map synchronized with second axial slice viewer 1402.
  • Second 3-D coronary vessel map 1404 may be presented in a top right area of second user interface 1400 and may include a 3-D grid to identify the length of the blood vessels in the blood vessel tree in each direction.
  • Second stretched blood vessel image 1406 includes a vertical bar which denotes a location of the slice displayed in second axial slice viewer 1402 in a stretched view of a selected blood vessel. Second stretched blood vessel image 1406 may be presented in a bottom left area of second user interface 1400.
  • Pathology report type view 1408 may contain a pathology list 1409 of detected pathologies based on the processes described with reference to Figs. 2-9.
  • the pathologies may include soft plaque, calcified plaque, and mixed plaque regions.
  • the location and stenosis level may be included for each pathology in pathology list 1409.
  • Selecting a pathology 1410 from pathology list 1409 of pathology report type view 1408 may cause a synchronized display of pathology 1410 in second axial slice viewer 1402, second 3-D coronary vessel map 1404, and second stretched blood vessel image 1406.
  • second axial slice viewer 1402 includes a first pathology indicator 1412, which indicates the location of pathology 1410 in second axial slice viewer 1402.
  • Second 3-D coronary vessel map 1404 includes a second pathology indicator 1414, which indicates the location of pathology 1410 in the 3-D coronary artery tree view.
  • Second stretched blood vessel image 1406 includes a first point 1416 and a second point 1418, which indicate the location of pathology 1410 in second stretched blood vessel image 1406.
  • a fifth user interface 1500 of visualization application 114 in accordance with a third exemplary embodiment is shown.
  • Fifth user interface 1500 may include a third axial slice viewer 1502 and a third stretched blood vessel image 1504.
  • Third axial slice viewer 1502 is synchronized with third stretched blood vessel image 1504.
  • Third axial slice 1502 presents intensity levels from an axial slice of imaging data. The intensity levels may be indicated in color or gray-scale.
  • third axial slice viewer 1502 may indicate an identified pathology 1506 in red.
  • Third stretched blood vessel image 1504 presents intensity levels of a blood vessel selected from third axial slice viewer 1502 and shown in stretched form.
  • Third stretched blood vessel image 1504 may includes a vertical bar 1508 which denotes a location of the slice presented in third axial slice viewer 1502.
  • third stretched blood vessel image 1504 shows a stretched presentation of the appropriate vessel.
  • third axial slice viewer 1502 displays the appropriate slice of the patient study.
  • fifth user interface 1500 may initially include third axial slice viewer 1502.
  • third axial slice viewer 1502 When the user selects an artery from third axial slice viewer 1502, the selected blood vessel is presented in third stretched blood vessel image 1504 with vertical bar 1508 denoting the location of the slice presented in third axial slice viewer 1502.
  • Execution of one or more of the processes described with reference to Figs. 2-9 may be performed after selection of the blood vessel to identify the stretched blood vessel presented in third stretched blood vessel image 1504.
  • the user may trigger a determination of all relevant segments of the blood vessel from the slices and reconstruct the stretched blood vessel for presentation in third stretched blood vessel image 1504.
  • Fifth user interface 1500 further may include an axial presentation only button 1510, a new study selection button 1512, a save current screen button 1514, and an exit program button 1516.
  • User selection of axial presentation only button 1510 causes stretched blood vessel image 1504 to be removed from fifth user interface 1500.
  • User selection of new study selection button 1512 causes presentation of a selection window that allows the user to select a new patient study for analysis.
  • User selection of save current screen button 1514 causes presentation of a save window that allows the user to select a location and a name for a file to which the contents of fifth user interface 1500 are saved for review, for printing, for sending with a message, etc.
  • User selection of exit program button 1516 may cause fifth user interface 1500 to close.
  • exemplary operations associated with assessing a quality of the medical imaging data are described. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. The operations described with reference to Fig. 16 may be performed before execution of pathology identification application 112 and/or visualization application 114 (described with reference to Fig. 1) or integrated as part of execution of pathology identification application 112 and/or visualization application 114. For example, the exemplary operations may be performed after operation 210 described with reference to Figs. 2a and 2b to determine a success or failure of the image enhancement process. In an operation 1600, an image resolution is calculated.
  • Image resolution may be stored in the image DICOM header and defined based on the "distance between slices" and "pixel size” fields.
  • the calculated image resolution is compared with a quality control parameter such as an image resolution threshold.
  • An exemplary image resolution threshold may be based on an inter-slice distance below 0.7 mm and a pixel size below 0.5 mm.
  • a determination is made concerning whether or not the image resolution is high enough. If the calculated image resolution is high enough, processing continues at an operation 1606. If the calculated image resolution is not high enough, processing continues at operation 1624.
  • a zoom in used in reconstruction should cause the heart region to occupy most of the region of interest (ROI).
  • the heart region may be automatically estimated and, if it is significantly smaller than the ROI, it may be determined that the heart region is not fully visible. If the heart region is fully visible, processing continues at an operation 1608. If the heart region is not fully visible, processing continues at operation 1624.
  • the volume of contrast may be estimated by thresholding the image data for example at 100 HU.
  • the determined volume of contrast is lower than a predefined threshold such as 30 mm x 30 mm x 20 mm, it may be determined that the contrast injection density is unacceptable. If the contrast injection density is acceptable, processing continues at an operation 1610. If the contrast injection density is not acceptable, processing continues at operation 1624.
  • the image may be searched for various artifacts using specialized filters.
  • operation 1610 a determination is made concerning whether or not motion artifacts exist. If motion artifacts do not exist, processing continues at an operation 1612. If motion artifacts exist, processing continues at operation 1624.
  • operation 1612 a determination is made concerning whether or not reconstruction artifacts exist. If reconstruction artifacts do not exist, processing continues at an operation 1614. If reconstruction artifacts exist, processing continues at operation 1624.
  • operation 1614 a determination is made concerning whether or not physical artifacts exist. If physical artifacts do not exist, processing continues at an operation 1616. If physical artifacts exist, processing continues at operation 1624.
  • an image signal-to-noise ratio is calculated.
  • the calculated SNR is compared with a quality control parameter such as a SNR threshold.
  • the image SNR may be computed as [(M n , -M IB )/ ⁇ m ] 2 where M w is the mean of the intensity in the vessel, M IB is the mean of the intensity of the background, and ⁇ IB is the standard deviation of the intensity of the background.
  • the vessel and background locations may be coarsely estimated in order to collect the statistics.
  • a determination is made concerning whether or not the calculated SNR is high enough.
  • a SNR threshold of 100 dB is used.
  • a message is output to the user.
  • a message may be printed from computing device 102, may be displayed at computing device 102, may played audibly from computing device 102, may be transmitted to the user (i.e., e-mail, text message, instant message, voice message, etc.), etc.
  • the message may indicate an assessment of the image data quality based on the test condition which failed.
  • processing of the CT image data stops.
  • One or more quality control parameters may be selected by the user for example using a user interface presented using pathology identification application 112 and/or visualization application 114.
  • the selected quality control parameters may be defined and/or stored in a configuration file stored at computing device 102 or on another computing device accessible from computing device 102.
  • the quality control parameters may be defined as a function of the type of medical diagnostic equipment used to generate the imaging data.
  • the quality control parameters may be predefined.
  • Comparison with quality control parameters allows supervision of the quality of the results produced by pathology identification application 112 and/or visualization application 114. A failure of a test comparison performed on a partial result may indicate either a total failure of the automatic processing or a low confidence in the result. In the former case, the processing should be stopped.
  • the confidence level should be propagated to the next processing level and finally presented to the user.
  • Each module of the processing pipeline in addition to the expected output generates a number from zero to one designating the module confidence in the result.
  • the confidence is computed differently and the computation depends on the particular algorithm.
  • the overall confidence of the system in the final results is computed as the product of confidence levels produced by each module.
  • a failure may be identified when a confidence level exceeds a threshold.
  • the failure in compliance with a required quality level of the input imaging data should prevent the system from automatic processing of the study and warn the operator about the low quality of the acquired data. The operator may decide in such case to perform a repeat examination.
  • the image quality control system can be used as a separate module (i.e. without pathology identification application 112 and/or visualization application 114) in a medical imaging facility as a tool for automatic acquisition quality control and CT machine calibration.

Abstract

A method and a system for identifying a data quality issue associated with medical diagnostic data is provided. The system includes, but is not limited to, an imaging apparatus configured to generate imaging data and a processor operably coupled to the imaging apparatus to receive the imaging data. The processor is configured to analyze the received imaging data; to identify a data quality issue associated with the received imaging data based on the analysis of the received imaging data; to generate a message including the identified data quality issue; and to send the generated message to a user.

Description

METHOD AND SYSTEM FOR AUTOMATIC QUALITY CONTROL USED IN COMPUTERIZED ANALYSIS OF CT ANGIOGRAPHY
FIELD
[0001] The field of the disclosure relates generally to computer systems. More specifically, the disclosure relates to automatic quality control used in computerized analysis of coronary CT angiography analysis.
BACKGROUND
[0002] Automatic computerized analysis of medical imaging data is becoming an increasingly important and acceptable method of working with vast amounts of data produced by modern medical diagnostic equipment. The reliability of the results produced by the automatic computerized analysis of the medical data is very important. The reliability and the robustness of automatic computerized analysis systems makes it either efficient or useless in a real medical world scenario.
[0003] No computer system can produce 100% correct results 100% of the time, which means that any automatic analysis system will fail eventually to analyze the data correctly. Such failures may be tolerable if the system automatically recognizes them and reports an inability to cope with the input to the operator. A failure report is more useful if the system provides additional details concerning the reason for the failure. For example, if the input data was of insufficient quality for automatic analysis, the input data is probably also of insufficient quality for a human to read as well. As a result, the medical test should be repeated.
[0004] In general, existing medical imaging data analysis tools do not provide a fully automatic solution and instead rely on the user's judgment concerning the quality of the result. As a result, these types of analysis tools assume a significant amount of user interaction, and thus, reserve the image understanding to the human operator. Another group of existing systems, which fall into the category of computer aided detection/diagnosis (CAD) tools, are usually more automatic and do include computerized image understanding, reasoning, and decision making. However, existing CAD systems lack a self-diagnosis capability, which could reduce the error rate by reporting cases where a reliable diagnosis is not possible and/or where the system is not confident in the validity of the result. Thus, a method and a system for an automatic self-diagnosis of medical CAD systems that allows automatic detection of cases in which automatic processing fails or produces unreliable results is needed. A method and a system for automatically quantifying an estimation of an image quality is also needed.
SUMMARY
[0005] A method and a system for automatic quality control (QC) for computer based analysis of medical imaging data is provided. In an exemplary embodiment, the medical imaging data is coronary computed tomography (CT) angiography (CTA) data. The method and system provide self-diagnosis of an automatic analysis process and provide assessment of a CT angiography study quality. The QC method and system automatically distinguish between cases for which a reliable result can be produced by the computerized analysis and those for which the automatic processing is impossible and/or a confidence in the result is low. In addition, the QC method and system provide an estimation of the image quality, which can be used independently by medical imaging equipment operators to measure the readability of the study and to calibrate the medical imaging equipment.
[0006] In an exemplary embodiment, a system for identifying a data quality issue associated with medical diagnostic data is provided. The system includes, but is not limited to, an imaging apparatus configured to generate imaging data and a processor operably coupled to the imaging apparatus to receive the imaging data. The processor is configured to analyze the received imaging data; to identify a data quality issue associated with the received imaging data based on the analysis of the received imaging data; to generate a message including the identified data quality issue; and to send the generated message to a user.
[0007] In another exemplary embodiment, a method of identifying a data quality issue associated with medical diagnostic data is provided. Imaging data is received at a computing device. The received imaging data is analyzed at the computing device. A data quality issue associated with the received imaging data is identified based on the analysis of the received imaging data at the computing device. A message including the identified data quality issue is generated at the computing device. The generated message is sent to a user.
[0008] In yet another exemplary embodiment, computer-readable instructions are provided that, upon execution by a processor, cause the processor to implement the operations of the method of identifying a data quality issue associated with medical diagnostic data
[0009] Other principal features and advantages of the invention will become apparent to those skilled in the art upon review of the following drawings, the detailed description, and the appended claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Exemplary embodiments will hereafter be described with reference to the accompanying drawings, wherein like numerals denote like elements.
[0011] Fig. 1 depicts a block diagram of an automated CT image processing system in accordance with an exemplary embodiment.
[0012] Figs. 2a and 2b depict a flow diagram illustrating exemplary operations performed by the automated CT image processing system of Fig. 1 in accordance with an exemplary embodiment.
[0013] Fig. 3 depicts a flow diagram illustrating exemplary operations performed in detecting a heart region in accordance with an exemplary embodiment.
[0014] Fig. 4 depicts a flow diagram illustrating exemplary operations performed in detecting a lung region in accordance with an exemplary embodiment.
[0015] Figs. 5a and 5b depict a flow diagram illustrating exemplary operations performed in detecting and identifying the aorta in accordance with an exemplary embodiment. [0016] Figs. 6a and 6b depict a flow diagram illustrating exemplary operations performed in identifying coronary artery vessel exit points from the aorta in accordance with an exemplary embodiment.
[0017] Figs. 7a and 7b depict a flow diagram illustrating exemplary operations performed in identifying coronary artery vessel tree in accordance with an exemplary embodiment.
[0018] Fig. 8 depicts a flow diagram illustrating exemplary operations performed in identifying calcium seed in accordance with an exemplary embodiment.
[0019] Fig. 9 depicts a flow diagram illustrating exemplary operations performed in identifying in identifying soft plaque in accordance with an exemplary embodiment.
[0020] Fig. 10 depicts a graph illustrating X-junction removal from a coronary artery vessel tree in accordance with an exemplary embodiment.
[0021] Fig. 11 depicts a first user interface of a visualization application presenting summary pathology results in accordance with an exemplary embodiment.
[0022] Fig. 12 depicts a second user interface of the visualization application presenting multiple views of a pathology in accordance with a first exemplary embodiment.
[0023] Fig. 13 depicts a third user interface of the visualization application presenting a blood vessel effective lumen area as a function of distance from the aorta in accordance with an exemplary embodiment.
[0024] Fig. 14 depicts a fourth user interface of the visualization application presenting multiple views of a pathology in accordance with a second exemplary embodiment.
[0025] Fig. 15 depicts a fifth user interface of the visualization application presenting multiple views of a pathology in accordance with a third exemplary embodiment.
[0026] Fig. 16 depicts a flow diagram illustrating exemplary operations performed in assessing a quality of medical imaging data in accordance with an exemplary embodiment. DETAILED DESCRIPTION
[0027] With reference to Fig. 1 , a block diagram of an image processing system 100 is shown in accordance with an exemplary embodiment. Image processing system 100 may include a computed tomography (CT) apparatus 101 and a computing device 102. Computing device 102 may include a display 104, an input interface 106, a memory 108, a processor 110, a pathology identification application 112, and a visualization application 114. In the embodiment illustrated in Fig. 1 , CT apparatus 101 generates image data. In general, however, the present techniques are well-suited for use with a wide variety of medical diagnostic system modalities, including magnetic resonance imaging systems, ultrasound systems, positron emission tomography systems, nuclear medicine systems, etc. Moreover, the various modality systems may be of a different type, manufacture, and model. Thus, different and additional components may be incorporated into computing device 102. Components of image processing system 100 may be positioned in a single location, a single facility, and/or may be remote from one another. As a result, computing device 102 may also include a communication interface, which provides an interface for receiving and transmitting data between devices using various protocols, transmission technologies, and media as known to those skilled in the art. The communication interface may support communication using various transmission media that may be wired or wireless.
[0028] Display 104 presents information to a user of computing device 102 as known to those skilled in the art. For example, display 104 may be a thin film transistor display, a light emitting diode display, a liquid crystal display, or any of a variety of different displays.
[0029] Input interface 106 provides an interface for receiving information from the user for entry into computing device 102 as known to those skilled in the art. Input interface 106 may use various input technologies including, but not limited to, a keyboard, a pen and touch screen, a mouse, a track ball, a touch screen, a keypad, one or more buttons, etc. to allow the user to enter information into computing device 102 or to make selections presented in a user interface displayed on display 104. Input interface 106 may provide both an input and an output interface. For example, a touch screen both allows user input and presents output to the user.
[0030] Memory 108 is an electronic holding place or storage for information so that the information can be accessed by processor 110 as known to those skilled in the art. Computing device 102 may have one or more memories that use the same or a different memory technology. Memory technologies include, but are not limited to, any type of RAM, any type of ROM, any type of flash memory, etc. Computing device 102 also may have one or more drives that support the loading of a memory media such as a compact disk or digital video disk.
[0031] Processor 110 executes instructions as known to those skilled in the art. The instructions may be carried out by a special purpose computer, logic circuits, or hardware circuits. Thus, processor 110 may be implemented in hardware, firmware, software, or any combination of these methods. The term "execution" is the process of running an application or the carrying out of the operation called for by an instruction. The instructions may be written using one or more programming language, scripting language, assembly language, etc. Processor 110 executes an instruction, meaning that it performs the operations called for by that instruction. Processor 110 operably couples with display 104, with input interface 106, with memory 108, and with the communication interface to receive, to send, and to process information. Processor 110 may retrieve a set of instructions from a permanent memory device and copy the instructions in an executable form to a temporary memory device that is generally some form of RAM. Computing device 102 may include a plurality of processors that use the same or a different processing technology.
[0032] Pathology identification application 112 performs operations associated with the analysis imaging data to identify and evaluate blood vessel structures and to identify and evaluate pathologies associated with the identified blood vessel structures. Some or all of the operations and interfaces subsequently described may be embodied in pathology identification application 112. The operations may be implemented using hardware, firmware, software, or any combination of these methods. With reference to the exemplary embodiment of Fig. 1 , pathology identification application 112 is implemented in software stored in memory 108 and accessible by processor 110 for execution of the instructions that embody the operations of pathology identification application 112. Pathology identification application 112 may be written using one or more programming languages, assembly languages, scripting languages, etc. Pathology identification application 112 may integrate with or otherwise interact with visualization application 114.
[0033] Visualization application 114 performs operations associated with presentation of the blood vessel analysis and identification results to a user. Some or all of the operations and interfaces subsequently described may be embodied in visualization application 114. The operations may be implemented using hardware, firmware, software, or any combination of these methods. With reference to the exemplary embodiment of Fig. 1 , visualization application 114 is implemented in software stored in memory 108 and accessible by processor 110 for execution of the instructions that embody the operations of visualization application 114. Visualization application 114 may be written using one or more programming languages, assembly languages, scripting languages, etc.
[0034] CT apparatus 101 and computing device 102 may be integrated into a single system such as a CT imaging machine. CT apparatus 101 and computing device 102 may be connected directly. For example, CT apparatus 101 may connect to computing device 102 using a cable for transmitting information between CT apparatus 101 and computing device 102. CT apparatus 101 may connect to computing device 102 using a network. In an exemplary embodiment, computing device 102 is connected to a hospital computer network and a picture archive, and communication system (PACS) and receives a CT study acquired on CT apparatus 101. Using PACS, CT images are stored electronically and accessed using computing device 102. CT apparatus 101 and computing device 102 may not be connected. Instead, the CT study acquired on CT apparatus 101 may be manually provided to computing device 102. For example, the CT study may be stored on electronic media such as a compact disc or a digital video disc. After receiving the CT study, computing device 102 may initiate automatic processing of the set of images that comprise the CT study. In an exemplary embodiment, CT apparatus 101 is a 64-slice multi-detector advanced CT scanner having a reconstructed slice width and inter-slice distance less than or equal to approximately 0.5 millimeters (mm), which produces standard digital imaging and communications in medicine (DICOM) images. Computing device 102 may be a computer of any form factor.
[0035] Image processing system 100 may provide an initial classification and decision support system, which allows fast and accurate ruling out of the three major diseases associated with chest pain. Image processing system 100 can be provided as a primary CT study inspection tool assisting an emergency room physician. Additionally, image processing system 100 can be used either to completely rule out some or all of the three diseases (in case of negative results) or as a trigger to call a radiologist and/or a cardiologist to further analyze the case. Additionally, image processing system 100 may automatically identify and segment blood vessel trees and automatically analyze each blood vessel to detect and map all relevant pathologies, including calcified and soft plaque lesions and their degree of stenosis.
[0036] With reference to Figs. 2a and 2b, exemplary operations associated with pathology identification application 112 and visualization application 114 of Fig. 1 are described. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. In an operation 200, pathology identification application 112 receives CT image data. The CT image data may be received from CT apparatus 101 directly or using a network. The CT image data also may be received using a memory medium. In an operation 202, a heart region is identified from the received CT image data. Data associated with the identified heart region is stored at computing device 102. In an exemplary embodiment, the data associated with the identified heart region includes a heart bounding box.
[0037] With reference to Fig. 3, exemplary operations associated with identifying the heart region data are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. For larger studies that may include head and neck or abdominal regions, a determination of the heart region provides a correct anatomical starting point for further segmentation. For smaller studies, a determination of the heart region reduces the size of the processed region and removes the non-relevant areas to reduce the false alarm risk. In an exemplary embodiment, the top and bottom boundaries of the heart region are cut a predetermined distance above and below the heart center.
[0038] In an operation 300, a first projection into an X-Y plane is defined. A positive X-axis is defined as extending out from the left side of the body. A positive Y-axis is defined as extending out from the back side of the body. A positive Z-axis is defined as extending out from the head of the body. The first projection is defined by summing the DICOM series along the Z-axis. In an operation 302, a threshold is applied to the first projection. For example, a threshold greater than approximately zero may be applied to eliminate the negative values of air which dominate the region outside the heart region and to retain the positive Hounsfeld unit (HU) values which include the fat, blood, and bones within the heart region. In an operation 304, a first largest connect component (CC) is identified in the thresholded first projection. In an operation 306, a first center of mass of the first largest CC is determined and denoted as Xc, Yci.
[0039] In an operation 308, a second projection into a Y-Z plane is defined. The second projection is defined by summing the DICOM series along the X-axis. In an operation 310, a threshold is applied to the second projection data. For example, a threshold greater than approximately zero may be applied to eliminate the negative values of air which dominate the region outside the heart region and to retain the positive HU values which include the fat, blood, and bones within the heart region. In an operation 312, a second largest CC is identified in the thresholded second projection data. In an operation 314, a second center of mass of the second largest CC is determined and denoted as YC2, Z0. A heart region center is defined as X0, Yd, Z0. In an operation 316, a heart region bounding box is defined from the heart region center and an average heart region width in each axis direction, Wx, Wγ, Wz.
[0040] In an operation 318, the defined heart region is compared to a quality control parameter associated with a valid heart region. For example, the size and the position of the defined heart region is compared with an expected size and position within the CT study. Additionally, a tissue type, a contrast, and arterials inside the defined heart region may be compared with the expected tissue type, contrast, and arterials inside a heart region. In an operation 320, a determination is made concerning whether or not the heart region data is valid. If the heart region data is valid, processing continues at an operation 322. If the heart region data is not valid, processing continues at an operation 324. In operation 322, the defined heart region bounding box is stored at computing device 102. The X-axis, Y-axis, Z-axis system centered at Xc, Yc1, Zc defines a body coordinate system. In operation 324, a message is output to the user. For example, a message may be printed from computing device 102, may be displayed at computing device 102, may played audibly from computing device 102, may be transmitted to the user (i.e., e-mail, text message, instant message, voice message, etc.), etc. The message may include an error indication that identifies a source of the invalidity or unreliability determination. For example, the message may indicate an invalidity of the heart region determination. In an operation 326, processing of the CT image data stops.
[0041] With reference again to Fig. 2, in an operation 204, a lung region is identified from the received CT image data. Data associated with the identified lung region is stored at computing device 102. In an exemplary embodiment, the data associated with the identified lung region may include a lung mask. With reference to Fig. 4, exemplary operations associated with identifying the lung region data are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting.
[0042] In an operation 400, a lung threshold is applied to each slice of the DICOM series data. For example, a lung threshold of -400 HU may be applied. In an operation 402, a morphological filter is applied to the binary image. In an exemplary embodiment, the binary image is filtered using a morphological closing operation to define a lung region in the CT image data. Other processes for filling holes in the image may be used as known to those skilled in the art. In an operation 404, a lung region is defined. In an operation 406, the defined lung region is compared to a quality control parameter associated with a valid lung region. For example, the intensity levels and patterns inside the defined lung region are compared with expected intensity levels and patterns in an lung region. In an operation 408, a determination is made concerning whether or not the lung region data is valid. If the lung region data is valid, processing continues at an operation 410. If the lung region data is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the lung region determination. In operation 410, the defined lung region is stored at computing device 102.
[0043] With reference again to Fig. 2, in an operation 206, a thorax region is identified from the received CT image data. Data associated with the identified thorax region is stored at computing device 102. The thorax region may be defined as a convex hull of the lungs and the diaphragm. In an operation 208, the pulmonary arteries are identified from the received CT image data. Data associated with the identified pulmonary arteries are stored at computing device 102.
[0044] In an operation 210, the received CT image data is preprocessed. For example, preprocessing may include image enhancement, smoothing, noise reduction, acquisition artifacts detection, etc. Examples of image enhancement algorithms include Gaussian smoothing, median filtering, bilateral filtering, anisotropic diffusion, etc. In an operation 211 , image quality test data is calculated to determine a quality of the image. In an operation 212, the calculated image quality test data is compared to quality control parameters associated with quality image data. In an operation 213, a determination is made concerning whether or not the image quality is satisfactory. An exemplary process for evaluating a quality of the image data is shown with reference to Fig. 16. If the image quality is satisfactory, processing continues at an operation 214. If the image quality is unsatisfactory, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an unsatisfactory image quality.
[0045] In an operation 214, the left and right main pulmonary artery trees are defined. In an operation 216, the lumen of the left and right main pulmonary artery trees is analyzed to identify any pulmonary embolism candidates. In an operation 218, any pulmonary embolism candidates are classified. In an operation 220, possible pulmonary embolism lesions are identified. [0046] In an operation 222, the ascending and the visible part of the abdominal aorta are segmented. In an operation 224, the lumen of the abdominal aorta is segmented. In an operation 226, a 3-D geometry of the abdominal aorta is modeled. In an operation 228, the modeled aorta is compared to a hypothesized "normal" abdominal aorta to identify deviations from the hypothesized "normal" abdominal aorta. In an operation 230, suspicious locations are detected and analyzed to identify dissections and aneurysms.
[0047] In an operation 232, the aorta is identified. The aorta is detected in the first imaging slice of the heart region bounding box based, for example, on intensity and shape properties including circularity, compactness, and area. The remainder of the aorta is identified by moving from slice to slice and looking for a similar 2-D object in each slice. Data associated with the identified aorta is stored at computing device 102. In an exemplary embodiment, the data associated with the identified aorta includes an aorta shape and boundary in the identified heart region. It is assumed that the heart region bounding box detected on the previous step includes the aorta exit from the heart and that the cross section of the aorta in the upper slice of the heart region is approximately circular.
[0048] With reference to Figs. 5a and 5b, exemplary operations associated with identifying the aorta are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. In an operation 500, a first slice is selected from the heart region data. In an operation 502, an aorta threshold is applied to the first slice of the DICOM series data. For example, a lung threshold of 200 HU may be used. In an operation 504, a morphological filter is applied to the binary image. In an exemplary embodiment, the binary image is filtered using a series of morphological filtering operators including an opening operator using a first parameter. In an operation 506, one or more CCs are identified.
[0049] In an operation 508, a compactness of each identified CC is determined. In an operation 510, identified CCs having a determined compactness that exceeds a compactness threshold are eliminated from further consideration. A compactness measure of a shape is a ratio of the area of the shape to the area of a circle (the most compact shape) having the same perimeter. The ratio may be expressed mathematically as M = 4w(area) / 2(perimeter). In an exemplary embodiment, the compactness threshold is 0.75. In an operation 512, a size of each identified connected component is determined. In an operation 514, identified CCs having a size that exceeds a maximum size threshold or that is below a minimum size threshold are eliminated from further consideration. In an exemplary embodiment, the maximum size threshold is 10,000 pixels. In an exemplary embodiment, the minimum size threshold is 1 ,000 pixels. In an operation 516, a determination is made concerning whether or not any identified CCs remain for consideration. If identified CCs remain for consideration, processing continues at an operation 518. If no identified CCs remain for consideration, processing continues at an operation 520. In operation 518, an initial aorta candidate is selected from the remaining CCs. For example, if a plurality of identified CCs remain for consideration, the largest candidate CC that is foremost in the body is selected as the initial aorta candidate.
[0050] In operation 520, a next slice is selected from the heart region data. In an operation 522, the aorta threshold is applied to the next slice of the DICOM series data. In an operation 524, the morphological filter is applied to the binary image. In an operation 526, one or more CCs are identified. In an operation 528, a compactness of each identified CC is determined. In an operation 530, the identified CCs having a determined compactness that exceeds the compactness threshold are eliminated from further consideration. In an operation 532, a size of each identified connected component is determined. In an operation 534, the identified CCs having a size that exceeds the maximum size threshold or that is below the minimum size threshold are eliminated from further consideration. In an operation 536, a determination is made concerning whether or not any identified CCs remain for consideration in the current slice. If identified CCs remain for consideration, processing continues at an operation 538. If no identified CCs remain for consideration, processing continues at operation 520.
[0051] In operation 538, the identified CCs from the current slice are compared with the aorta candidate object(s) created from the previous slice(s). In an operation 540, a determination is made concerning whether or not any identified CCs match CCs identified from the previous slices. In an operation 542, if a match is found between a CC and an aorta candidate object, the matched CC is assigned to the aorta candidate object. For example, if a center of a CC is closer than twenty pixels to the center of an aorta candidate object, the CC may be identified as matched with the aorta candidate object. In an operation 544, if a match is not found between a CC and an aorta candidate object, a new aorta candidate object is created based on the CC.
[0052] In an operation 546, a determination is made concerning whether or not all of the slices have been processed. If slices remain, processing continues at operation 520. If no slices remain, in an operation 548, aorta candidate objects are eliminated based on length. For example, aorta candidate objects that persist for less than 20 slices may be removed from further consideration. In an operation 550, an aorta object is selected from the remaining aorta candidate objects. For example, the aorta candidate object closest to the upper left corner of the image may be selected as the aorta object. In an operation 552, a bounding box is defined around the selected aorta object to identify a region in which the aorta is located in the CT image data. In an operation 554, test data associated with the selected aorta object are calculated. For example, geometrical parameters associated with the selected aorta object and its relative position inside the heart bounding box are measured and/or calculated as test data. In an operation 556, the calculated test data are compared to a quality control parameter associated with a valid aorta object. For example, the diameter of the aorta should not change too much from the aortic arch to the entry point to the heart. Additionally, aorta cross sections should be approximately circular with a diameter within a valid range such as from 20 mm to 50 mm. Some or all of these parameters may be used as quality control parameters associated with determining if a selected aorta object is valid. In an operation 558, a determination is made concerning whether or not the selected aorta object is valid. If the selected aorta object is valid, processing continues at an operation 560. If the selected aorta object is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the aorta object determination. In operation 560, the selected aorta object is stored at computing device 102.
[0053] With reference again to Fig. 2, in an operation 234, exit points of the coronary arteries from the aorta are identified by evaluating all structures connected to the aorta object which look like a vessel. The direction of the vessel near a link point with the aorta object should be roughly perpendicular to an aorta centerline. Additionally, the left and right coronary arteries are expected to exit from the aorta object in a certain direction relative to the body coordinate system. If there are several exit point candidates, the exit point candidate which leads to a larger blood vessel tree is selected. It is assumed that the selected aorta object includes the points where the left and right coronary trees connect with the aorta.
[0054] With reference to Fig. 6, exemplary operations associated with identifying the exit points of the coronary arteries from the aorta object are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. Imaging slices including the aorta bounding box and a mask of the aorta detected at a previous slice are processed to detect the aorta at the current slice. In an operation 600, a first slice is selected from the aorta object. In an operation 602, regions are segmented based on a segmentation threshold at the detected aorta edges. The segmentation threshold may be calculated from the median value of pixels of the smoothed image at the detected edges. A ring of pre-defined radius is defined around the aorta edges detected on the previous slice and the edges are found in the ring on the current slice. Small edges are removed from further consideration. The segmentation threshold may be selected adaptively. For example, if no edges are found in the ring using the calculated segmentation threshold, the calculated segmentation threshold is reduced by half, and the procedure is repeated. If no edges are found using the lowered segmentation threshold, the calculated segmentation threshold is used for subsequent slices.
[0055] In an operation 604, the segmented image is post-processed. For example, small segmented objects are removed, possible vessels are removed from the segmented aorta candidates, and the segmented aorta candidates are intersected with the aorta detected in the previous slice. In an operation 606, the aorta candidates are validated by ensuring that there is at least one candidate that intersected the aorta detected in the previous slice and by ensuring that the aorta does not grow too fast. For example, if the aorta size in both a previous and a current slice is larger than 1500 pixels, the size growth ratio may be limited to 1.4. In an operation 608, the aorta candidates are selected. For example, CCs with a small intersection with the previously detected aorta are removed from consideration, and the upper-left-most candidate is chosen if a plurality of aorta candidates exist in the current slice. In an operation 610, the compactness of the selected aorta candidate is checked to ensure that the candidate is not compact. If the aorta candidate is not compact, the aorta search window is limited for the next slice. If the aorta candidate is compact, the whole image is used to search for the aorta in the next slice. In an operation 612, a bounding box for the aorta is calculated. If the aorta candidate is not compact, the bounding box size may be fixed and only the position of the bounding box updated to compensate for aorta movement. If the aorta candidate is compact, the bounding box may be attached to the upper left side of the aorta.
[0056] In an operation 614, a vesselness score is calculated for each voxel of the aorta object. As known to those skilled in the art, the vesselness score can be determined using a vesselness function. A vesselness function is a widely used function based on the analysis of Hessian eigen values. A good description of an exemplary vesselness function can be found for example in Frangi, A. F., Niessen, W. J., Vincken, K. L. and Viergever, M. A., 1998, "Multiscale Vessel Enhancement Filtering", MICCAI'98, LNCS 1496, pp. 130-137. In an operation 616, a vesselness threshold is applied to the calculated vesselness score to identify possible exit points based on the HU value of a ring around the aorta object. Pixels in a ring around the detected aorta are grouped into CCs, which are analyzed to choose the most probable candidates for coronary tree exit points. In an operation 618, possible exit points are filtered to remove false candidates. For example, the possible exit points may be filtered based on a size of the CC corresponding to the possible exit, a location of the CC relative to the aorta, an incident angle of the CC relative to the aorta, etc. In an operation 620, a determination is made concerning whether or not any exit points are left. If no exit points are left for this slice, processing continues at an operation 624. If one or more exit points are left for this slice, processing continues at an operation 622. In operation 622, the one or more exit points left for this slice are added to a possible exit points list. In an operation 624, a determination is made whether or not the last slice has been processed. If the last slice has not been processed, processing continues at an operation 626. In operation 626, the next slice is selected from the aorta object data and processing continues at operation 602.
[0057] If the last slice has been processed, processing continues at an operation 628. In operation 628, a CC is identified for each exit point included in the possible exit points list. In an operation 630, a volume is calculated for each exit point CC (EPCC). In an operation 632, any EPCC having a volume below a volume threshold is eliminated from further consideration as an exit point. For example, the volume threshold may be 1500 voxels. In an operation 634, a maximum width of each EPCC is calculated. In an operation 636, any EPCC having a maximum width below a width threshold is eliminated from further consideration as an exit point. For example, the width threshold may be 2 mm. In an operation 638, a distance to the aorta is calculated for each EPCC. In an operation 640, the EPCC having a minimum distance to the aorta is selected. In an operation 642, a determination is made concerning whether or not a plurality of EPCCs remain. If a plurality of EPCCs remain, processing continues at an operation 644. If a plurality of EPCCs do not remain, processing continues at an operation 646. In operation 644, the EPCC having a maximum width is selected from the plurality of EPCCs remaining. In operation 646, the exit point is identified from the selected EPCCs.
[0058] In an operation 648, the geometry of the selected EPCC including a change of the vessel radius and the position and orientation of the selected EPCC in the heart bounding box are analyzed. In an operation 650, the analysis data are compared to a quality control parameter associated with a valid EPCC object. For example, a valid EPCC object should extend for a certain minimal distance from the exit point. An exemplary minimum distance is two centimeters. Additionally, if the EPCC object is indeed a tree of vessels, its convex hull to volume ratio also should be relatively low. The width of the structures in the EPCC object also should be low compared to a maximal expected radius of the blood vessel such as three mm. The EPCC corresponding to the right coronary tree should approximately cover the statistically expected right coronary tree region. The same should be true for the left coronary tree. Some or all of these parameters may be used as quality control parameters associated with determining if a selected EPCC object is valid. In an operation 652, a determination is made concerning whether or not the selected EPCC object is valid. If the selected EPCC object is valid, processing continues at an operation 654. If the selected EPCC object is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the EPCC determination.
[0059] With reference again to Fig. 2, in an operation 236, a coronary artery vessel tree is defined. For a traversed section of the blood vessel tree, a set of end points is identified, and an attempt is made to continue tracking beyond the end point in the direction of the corresponding tree branch. If an additional tree segment is detected, it is connected to the traversed tree, and the processing continues recursively. The process is finished when no branch can be continued. The stopping condition may result in connecting a wrong structure to the coronary tree (e.g. a vein or some debris in a noisy CT study). As a result, the successfully tracked vessels are identified and those which remain to be detected are identified. For example, which blood vessels are to be segmented (e.g. the left coronary artery (LCA), the right coronary artery (RCA), the left main coronary artery (LM), the middle portion of the left anterior descending coronary artery (LAD), and the proximal portion of the left circumflex coronary artery (LCX) and others) may be defined as an input to the process. Additionally, a maximum vessel length to track (e.g. five centimeters from the aorta) and a minimum blood vessel diameter to continue tracking also may be defined as inputs to the process. The location of some blood vessels may be based on anatomical landmarks. For example, the RCA is located in the right atrioventricular plane. These anatomical landmarks, which may be collated through an anatomical priors processing operation, allow false structures in the "wrong" places to be discarded and support the location of lost branches in the "right" places. A graph representation of the segmented vessel tree can be built from the identified end points and bifurcation points. Graph nodes are the end points and the bifurcation points. The edges are segments of a vessel centerline between the nodes.
[0060] With reference to Figs. 7a and 7b, exemplary operations associated with defining the coronary artery vessel tree are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. In an operation 700, the vesselness score data is received. In an operation 702, a first binary volume is defined for a first threshold. The first binary volume includes a '1' for each voxel that exceeds the first threshold and a '0' for each voxel that does not exceed the first threshold. In an operation 704, a second binary volume is defined for a second threshold. The second binary volume includes a '1' for each voxel that exceeds the second threshold and a '0' for each voxel that does not exceed the second threshold. The second threshold has a higher HU value than the first threshold. In an operation 706, one or more CCs in the first binary volume that intersect voxels from the second binary volume are selected as the one or more vessel CCs (VCCs). In an exemplary embodiment, intersection may be determined based on a spatial proximity between the CCs. In an exemplary embodiment, the first threshold and the second threshold are selected based on a statistical analysis of the input data such that the amount of voxels above the first threshold is approximately 0.15% of the total number of voxels in the volume and such that the amount of voxels above the second threshold is approximately 0.45% of the total number of voxels.
[0061] In an operation 708, the selected VCCs are labeled in the first binary volume. In an operation 710, a starting point or root is selected for a first VCC. In an operation 712, a width map is calculated for the first VCC. The width map includes the width of the VCC or the inverse distance from any point in the VCC to the boundary of the VCC. Thus, small values are near the centerline of the VCC and larger values are at the edges of the VCC. In an operation 714, the exit point of the selected VCC is identified in the binary volume. The RCA tree has a single exit point. Additionally, the LM tree has a single exit point. In an operation 715, the exit point of the selected VCC is compared to a quality control parameter associated with a valid exit point. For example, a relative position along the aorta of the exit point of the selected VCC and a size of the tubular structure components connected to the selected VCC are compared with expected values. Additionally, statistics on the positions of certain blood vessels relative to the heart related coordinate system may be collected over a large database of studies. Any given VCC may be compared to the derived statistically expected location of the blood vessels and the likelihood of the VCC being that vessel is computed. Some or all of these parameters may be used as quality control parameters associated with determining if a selected exit point is valid. In an operation 716, a determination is made concerning whether or not the exit point is valid. If the exit point is valid, processing continues at an operation 717. If the exit point is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the exit point determination.
[0062] In an operation 717, a distance map is calculated for the first VCC. The distance is calculated from any point in the VCC to the identified exit point. The distance map includes the calculated distance weighted by the width to ensure that the minimal path follows the vessel centerline. In an operation 718, candidate endpoints are identified. For example, during the calculation of the weighted distance map, one or more candidate endpoints may be saved. The candidate endpoints are voxels, which did not update any of their neighbors during the distance map calculation. In an operation 720, non-local maxima candidate endpoints are filtered. Thus, the candidate endpoints are scanned and only local maxima with respect to the distance from the root over a given window are kept. This process eliminates candidates that are not true blob vessel end points.
[0063] In an operation 722, an identifier for the candidate endpoint is created. In an operation 724, a new vertex is added to a symbolic graph of the vessel tree. In an operation 726, a first candidate endpoint is backtracked to the root to create graph edges. An auxiliary volume is used to mark voxels that have already been visited. The back-tracking may be a gradient descent iterative process (the gradient is in the distance field). Because the weighted distance map contains a single global minimum (the root), convergence is guaranteed. The method used to define the distance map ensures that the backtracking will be along the centerline or close to it. In an operation 728, all visited voxels in the auxiliary volume are marked with the current vertex identifier during the backtracking.
[0064] In an operation 730, a determination is made concerning whether or not a root is reached. If a root is reached, processing continues at an operation 732. If a root is not reached, processing continues at an operation 734. In operation 732, a new edge and vessel path are defined based on the backtracking. Processing continues at an operation 740. In an operation 734, a determination is made concerning whether or not an already visited voxel is reached. If an already visited voxel is reached, processing continues at an operation 736. If an already visited voxel is not reached, processing continues at an operation 726 to continue the backtracking to the endpoint. In an operation 736, a new edge and a new bifurcation vertex are defined. In an operation 738, the new bifurcation vertex is connected to the currently backtracked path, and the new edge and the new bifurcation vertex are added to the vessel tree.
[0065] In an operation 742, a determination is made concerning whether or not the endpoint is a leaf of the vessel tree or a vessel disconnected due to a low vesselness measure. In an exemplary embodiment, the determination is made based on the direction of the vessel at the endpoint and by searching for another VCC in a vacancy that contains a nearby endpoint. If the endpoint is a leaf, processing continues at operation 722 to create a new graph. The two graphs are joined together. If the endpoint is not a leaf, processing continues at an operation 744. In an operation 744, a determination is made concerning whether or not the last endpoint has been processed. If the last endpoint has not been processed, processing continues at operation 722.
[0066] If the last endpoint has been processed, processing continues at an operation 746. In operation 746, short branches are removed based on the rationale that they do not contribute to the analysis because important findings are usually located at the major blood vessels, which are thick and elongated. Therefore, in an exemplary embodiment, graph edges which lead to endpoints that are less than a length threshold are removed. An exemplary length threshold is five mm. In an operation 748, x-junctions are removed to eliminate veins. Veins are usually faint and spatially close to the arteries. X-junctions are defined as two very close bifurcation points. For example, close bifurcation points may be less than approximately 3 mm from each other. In an exemplary embodiment, a bifurcation may be two VCCs intersecting at angles between approximately 70 degrees and approximately 110 degrees. Additionally, a bifurcation may be two VCCs intersecting at angles approximately equal to 90 degrees. The sub-tree which remains is the one which has the closest direction to the edge arriving from the aorta.
[0067] For example, with reference to Fig. 10, a vessel tree 1000 includes a first vessel path 1002, a second vessel path 1004, and a third vessel path 1006. First vessel path 1002 and second vessel path 1004 form a first x-junction 1008. First vessel path 1002 has the closest direction to the edge arriving from the aorta and is selected to remain in the vessel tree. Second vessel path 1004 is removed. First vessel path 1002 and third vessel path 1004 form a second x-junction 1010. Again, first vessel path 1002 has the closest direction to the edge arriving from the aorta and is selected to remain in the vessel tree. Third vessel path 1006 is removed.
[0068] With reference again to Figs. 7a and 7b, in an operation 750, single entry, single exit point vertices are removed. These vertices are created when one of the endpoints is recursively continued. The vertex is removed, and the two edges are joined to a single path. In an operation 752, a determination is made concerning whether or not all VCCs have been processed. If all VCCs have been processed, processing continues at an operation 754. If all VCCs have not been processed, processing continues at operation 710. In operation 754, blood vessel test data are calculated. For example, a volume and a length of the detected blood vessels is calculated. In an operation 756, the calculated blood vessel test data are compared to a quality control parameter associated with a valid blood vessel tree. For example, the presence of blood vessels in the expected anatomical locations such as the atrioventricular grooves and inter-ventricular plane is evaluated. The RCA is expected to go through the atrioventricular groove between the right ventricle and the right atrium. The groove may be automatically detected, and the RCA candidate vessel's location compared to the groove mask. If a significant part of the vessel is located outside the groove, or, alternatively, if a large part of the groove is not covered by the vessel, the vessel may be determined not to be the RCA. In an operation 758, a determination is made concerning whether or not the blood vessel tree is valid. If the blood vessel tree is valid, processing continues at an operation 760. In operation 760, processing continues. If the exit point is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the blood vessel tree determination.
[0069] With reference again to Fig. 2, in an operation 238, the defined coronary artery vessel tree is labeled. A list of graph edges (vessel segments) may be assigned to each blood vessel tracked by analyzing the relative section positions and locations relative to detected anatomical heart landmarks. In an operation 240, blood vessel test data are analyzed to determine a validity of the coronary artery tree labeling process. For example, the presence of the major coronary arteries such as LCA, RCA, LM, LAD, and LCX and their relative positions are analyzed. In an operation 242, the analyzed data is compared to a quality control parameter associated with valid blood vessel positions. For example, it is expected that the RCA and the LCX are approximately in one plane (RCA-LCX plane) and that the remaining blood vessels are on one side of the RCA-LCX plane. It further is expected that the LAD is approximately perpendicular to the RCA-LCX plane. In an operation 244, a determination is made concerning whether or not the blood vessel labeling process is valid. If the blood vessel labeling process is valid, processing continues at an operation 246. If the blood vessel labeling process is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the blood vessel labeling process. The blood vessel tree represented by a centerline for each blood vessel segment is stored at computing device 102.
[0070] In an operation 246, a radius of each blood vessel is determined. In an operation 248, a blood vessel centerline is determined. "Sausages" of blood vessels are obtained from the coronary artery vessel tree. Each "sausage" includes axial blood vessel cross-sections taken perpendicular to the blood vessel direction. Initially, a blood vessel center is presumed to be at the center of each section. Either "stretched" or "curved" blood vessels can be used. Any blood vessel radius and center line estimation method can be used. In an exemplary embodiment, low pass post-filtering between consecutive cross-sections is performed. Because a blood vessel may be surrounded by tissue having similar attenuation values, indirect indicators may be used to define the blood vessel edge. Areas having low values, which clearly don't belong to a blood vessel are identified, and the largest circle that lies outside the identified areas is defined. In an alternative embodiment, a largest circle that can be defined that fits into the valid (bright) area is defined. An arbitration process may be used to determine which approach should be used for each blood vessel. A blood vessel center consisting of a number of pixels can be defined, in particular when a cross section is elongated. In an exemplary embodiment, the blood vessel center is reduced to a single pixel.
[0071] In an operation 250, areas of calcium are identified in each blood vessel. Any high precision calcium identification method can be used. In an exemplary embodiment, a cross section based analysis aimed at location of the calcium seeds is performed, and a volume based analysis aimed at removal of spurious seeds created by the "salt noise" and by the growing of valid seeds into the correct calcium area is performed. With reference to Fig. 8, exemplary operations associated with identifying the areas of calcium, if any, in each blood vessel are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. In an operation 800, a first slice is selected from the heart region data. In an operation 802, a calcium threshold is applied to the first slice. In an exemplary embodiment, the calcium threshold is 150 intensity levels above the blood vessel lumen level. Adaptive threshold values taking into account expected lumen values are used. In an operation 804, a morphological filter is applied to the thresholded first slice based on the non-concentric nature of the calcium deposits. Empirical observations indicate that calcium tends to appear close to the blood vessel borders. [0072] In an operation 806, a maximum intensity in a given cross-section is identified as a possible location of a calcium seed. In an operation 808, a distance from the center to the maximum intensity is calculated. In an operation 810, a determination is made concerning whether or not the calculated distance exceeds a calcium distance threshold. The calcium distance threshold is based on a comparison with an estimated radius value. If the distance does not exceed the calcium distance threshold, processing continues in an operation 814. If the distance does exceed the calcium distance threshold, processing continues in an operation 812. In operation 812, an area of the calcium seed is calculated. In operation 814, a determination is made concerning whether or not any vessels remain for processing. If vessels remain, processing continues at operation 808. If no vessels remain, processing continues at an operation 816. In operation 816, a determination concerning whether or not any slices remain for processing is performed. If no slices remain, processing continues at an operation 820. If slices remain, processing continues at an operation 818. In operation 818, the next slice is selected from the heart region data and processing continues at operation 802. In operation 820, a volume of any identified calcium seed(s) is calculated based on the area calculated for each slice and the number of slices over which the identified calcium seed(s) extends. If a calcium seed is identified, it also is extended to the surrounding high intensity areas providing that no "spill to the center" occurs. An extent of the calcium seed may be determined based on a threshold. For example, lumen intensities exceeding approximately 650 HU may be considered to be calcified plaque or part of the calcium seed.
[0073] With reference again to Fig. 2, in an operation 252, areas of soft plaque are identified in each blood vessel by the low intensity inside the blood vessel area. Any high precision soft plaque identification method can be used. With reference to Fig. 9, exemplary operations associated with identifying the areas of soft plaque, if any, in each blood vessel are described in accordance with an exemplary embodiment. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. In an operation 900, a first slice is selected from the heart region data. In an operation 902, a soft plaque threshold is applied to the first slice. In an exemplary embodiment, the soft plaque threshold is between approximately 50 HU and approximately 200 HU. Adaptive threshold values taking into account expected lumen values are used in an exemplary embodiment. In an operation 904, a determination is made concerning whether or not calcium is present. The presence of calcium may indicate the presence of the frequent figure "8" shaped pattern. In the figure "8" shaped pattern, calcium is located in one of the ovals of the "8". A lumen is located in the other oval of the "8", and soft plaque connects the two ovals. If calcium is present, processing continues at an operation 906. If calcium is not present, processing continues at an operation 908. In operation 906, a soft plaque area is identified and processing continues at an operation 911.
[0074] In operation 908, a determination is made concerning whether or not a half-moon structure is located in the blood vessel lumen. If a half-moon structure is identified from the determination, processing continues at an operation 910. If a half- moon structure is not identified from the determination, processing continue at operation 911. In operation 910, a soft plaque area is identified. In operation 911 , an area of the identified soft plaque is calculated. In operation 912, a determination concerning whether or not any slices remain for processing is performed. If no slices remain, processing continues at an operation 916. If slices remain, processing continues at an operation 914. In operation 914, the next slice is selected from the heart region data, and processing continues at operation 902. In operation 916, a volume of any identified soft plaque area(s) is calculated based on the area calculated for each slice and the number of slices over which the identified soft plaque area(s) extends. In an operation 918, a volume of any identified calcium seed(s) is updated to include areas between the calcium seed and the blood vessel border and between the calcium and soft plaque areas to compensate for natural intensity low passing that may have occurred during the CT image acquisition.
[0075] With reference again to Fig. 2, in an operation 254, a severity of any obstructions identified containing soft plaque or calcium is calculated. Any obstruction computation method can be used. In an exemplary embodiment, a total obstruction ratio is calculated as a ratio of the total calcium and soft plaque areas divided by the total blood vessel area excluding the border area. In an exemplary embodiment, an obstruction is identified to be severe if the total obstruction ratio exceeds 50% for at least two consecutive cross sections. An examining physician may be allowed to control the threshold to achieve a system sensitivity matching their clinical requirements.
[0076] In some pathological cases, the cross section images may appear reasonably normal. In these cases, pathology must be identified based on the analysis of global variations. In an operation 256, global filters are applied to identify pathologies. For example, a first filter may be applied to identify a rapid decrease in the blood vessel radius. A second filter may be applied to identify a rapid decrease in the lumen intensity. A third filter may be applied to identify a rapid increase in the lumen intensity. The decisions from the series of filters may be cumulative. As a result, it is sufficient if a pathology is identified through use of one of the three filters. The filters may use the values of blood vessel radius and luminance as computed above. Use of the global filters takes into account that even healthy vessels feature significant radius and luminance variations in particular due to natural narrowing of the blood vessels, rapid changes in the vicinity of bifurcations (especially after the bifurcations), noise (in particular for relatively narrow vessels), etc. Anomalies identified by the global filters are discarded, if located in the vicinity of any bifurcations.
[0077] In an operation 258, pathology test data are calculated to determine a validity of the pathology identification process. For example, the number of identified pathologies per blood vessel segment is calculated. In an operation 260, the calculated data is compared to a quality control parameter associated with valid pathology test data. For example, a statistically expected number of detected pathologies per blood vessel segment for a specific groups of patients is determined and compared with the calculated pathology test data. In an operation 262, a determination is made concerning whether or not the pathology identification process is valid. For example, a significant deviation from the statistically expected number of detected pathologies per blood vessel segment for a specific groups of patients indicates invalidity. If the pathology identification process is valid, processing continues at an operation 264. If the pathology identification process is not valid, processing continues at operation 324 shown with reference to Fig. 3 wherein the message indicates an invalidity of the pathology identification process.
[0078] The operations described with reference to Figs. 2-9 have been applied to a set of 50 clinical patient studies. The same studies were analyzed by expert radiologists. Overall, 200 blood vessels were analyzed. Out of this benchmark, 42 cases were identified as having severe pathologies. The remaining 158 cases were deemed to have no pathologies or only moderate pathologies. Pathology identification application 112 identified all of the severe cases correctly with a false alarm rate of 11%.
[0079] With reference again to Fig. 2, in an operation 264, a summary report is provided to a user based on the processes described with reference to Figs. 2-9. With reference to Fig. 11 , a first user interface 1100 of visualization application 114 is shown. First user interface 1100 may include a header portion 1102. Header portion 1102 may include patient data, study data, and/or acquisition data. For example, data displayed in header portion 1102 may be obtained from a header of the DICOM data. First user interface 1100 further may include a blood vessel list portion 1104. Blood vessel list portion 1104 may include a list of the blood vessels in the created blood vessel tree. Displayed next to a name identifying each blood vessel may be information related to each blood vessel including a lumen status, a total number of lesions, a number of calcium lesions, and a number of soft plaque lesions. The lumen status may indicate "normal" or a percentage of blockage that may be a percentage range. If a plurality of lesions are present, the range may indicate the maximum blockage range. A maximum volume and Agatston score may be displayed for the calcium lesions. A maximum volume and score also may be displayed for the soft plaque lesions.
[0080] User selection of a blood vessel 1106 in blood vessel list portion 1104 may cause display of a detailed description of the lesions associated with the selected blood vessel in a detail portion 1108. Detail portion 1108 may include a list of the lesions. For each lesion, a segment name, a lesion type, a degree of stenosis value, a volume, a distance from the aorta, a distance from the blood vessel origin, an eccentricity, a degree of positive remodeling, and a morphological regularity may be shown. First user interface 1100 further may include a totals portion 1110. Totals portion 1110 may include summary data associated with a degree of stenosis, lesions, the number of stents, etc.
[0081] With reference again to Fig. 2, in an operation 266, a visualization of the blood vessels is provided to a user based on the processes described with reference to Figs. 2-9. With reference to Fig. 12, a second user interface 1200 of visualization application 114 in accordance with a first exemplary embodiment is shown. In the exemplary embodiment of Fig. 12, four simultaneous views of the same pathology may be shown to facilitate a correct diagnosis with each view presented in a different area of second user interface 1200. Each view may be created using a variety of graphical user interface techniques in a common window, in separate windows, or in any combination of windows. Second user interface 1200 may include a first axial slice viewer 1202, a first 3-D coronary vessel map 1204, a first stretched blood vessel image 1206, and an intra-vascular ultrasound (IVUS) type view 1208. First axial slice viewer 1202 presents intensity levels from a slice of imaging data. The intensity levels may be indicated in color or gray-scale. For example, first axial slice viewer 1202 may indicate an identified pathology 1203 in red. First axial slice viewer 1202 may present the slice of imaging data in a top left area of second user interface 1200.
[0082] First 3-D coronary vessel map 1204 provides a view of the blood vessel tree synchronized with first axial slice viewer 1202 to indicate the identified pathology 1203. First 3-D coronary vessel map 1204 may be presented in a top right area of second user interface 1200 and may include a 3-D grid to identify the length of the blood vessels in the blood vessel tree in each direction. Selecting an area of first stretched blood vessel image 1206 may cause image rotation of first 3-D coronary vessel map 1204 around its axis to facilitate a correct 3-D perception of the blood vessel structure. First 3-D coronary vessel map 1204 may be synchronized with first axial slice viewer 1202 to distinguish the selected blood vessel from the remaining blood vessels in the blood vessel tree. First 3-D coronary vessel map 1204 may be rotated using an input interface as known to those skilled in the art. Indicators may be provided in first 3-D coronary vessel map 1204 to indicate end points and bifurcations. For example, end points may be indicated using green circles and bifurcations may be indicated using red circles.
[0083] First stretched blood vessel image 1206 includes a vertical bar which denotes a location of the slice displayed in first axial slice viewer 1202 in a stretched view of a selected blood vessel. First stretched blood vessel image 1206 may be located in a bottom left area of second user interface 1200. The physician can superimpose corresponding plaque areas. For example, soft plaque may be indicated in red and calcified plaque indicated in blue. If desired, the physician can invoke an edit mode and correct automatic results.
[0084] With reference to Fig. 13, a third user interface of visualization application 114 graphically displays a lumen area of each blood vessel to clearly identify all stenosis lesions and to allow an evaluation of their severity. The graphical display includes a distance from the aorta on the X-axis and a lumen area on the Y-axis. A first curve 1300 indicates a normal blood vessel lumen. A second curve 1302 indicates a stenosis due to calcified plaque. A third curve 1304 indicates a stenosis due to soft plaque.
[0085] With reference to Fig. 14, a fourth user interface 1400 of visualization application 114 in accordance with a second exemplary embodiment is shown. Fourth user interface 1400 may include a second axial slice viewer 1402, a second 3-D coronary vessel map 1404, a second stretched blood vessel image 1406, and a pathology report type view 1408. Second axial slice viewer 1402 may include an axial slice of the imaging data presented in a top left area of second user interface 1400. provides a current location on the 3-D coronary vessel map synchronized with second axial slice viewer 1402. Second 3-D coronary vessel map 1404 may be presented in a top right area of second user interface 1400 and may include a 3-D grid to identify the length of the blood vessels in the blood vessel tree in each direction. Selecting an area of second 3-D coronary vessel map 1404 may cause image rotation around its axis facilitating a correct 3-D perception of the blood vessel structure. [0086] Second stretched blood vessel image 1406 includes a vertical bar which denotes a location of the slice displayed in second axial slice viewer 1402 in a stretched view of a selected blood vessel. Second stretched blood vessel image 1406 may be presented in a bottom left area of second user interface 1400.
[0087] Pathology report type view 1408 may contain a pathology list 1409 of detected pathologies based on the processes described with reference to Figs. 2-9. The pathologies may include soft plaque, calcified plaque, and mixed plaque regions. The location and stenosis level may be included for each pathology in pathology list 1409. Selecting a pathology 1410 from pathology list 1409 of pathology report type view 1408 may cause a synchronized display of pathology 1410 in second axial slice viewer 1402, second 3-D coronary vessel map 1404, and second stretched blood vessel image 1406. For example, second axial slice viewer 1402 includes a first pathology indicator 1412, which indicates the location of pathology 1410 in second axial slice viewer 1402. Second 3-D coronary vessel map 1404 includes a second pathology indicator 1414, which indicates the location of pathology 1410 in the 3-D coronary artery tree view. Second stretched blood vessel image 1406 includes a first point 1416 and a second point 1418, which indicate the location of pathology 1410 in second stretched blood vessel image 1406.
[0088] With reference to Fig. 15, a fifth user interface 1500 of visualization application 114 in accordance with a third exemplary embodiment is shown. Fifth user interface 1500 may include a third axial slice viewer 1502 and a third stretched blood vessel image 1504. Third axial slice viewer 1502 is synchronized with third stretched blood vessel image 1504. Third axial slice 1502 presents intensity levels from an axial slice of imaging data. The intensity levels may be indicated in color or gray-scale. For example, third axial slice viewer 1502 may indicate an identified pathology 1506 in red. Third stretched blood vessel image 1504 presents intensity levels of a blood vessel selected from third axial slice viewer 1502 and shown in stretched form. Third stretched blood vessel image 1504 may includes a vertical bar 1508 which denotes a location of the slice presented in third axial slice viewer 1502. When the user selects a vessel area in third axial slice viewer 1502, third stretched blood vessel image 1504 shows a stretched presentation of the appropriate vessel. When the user selects an area in third stretched blood vessel image 1504, third axial slice viewer 1502 displays the appropriate slice of the patient study.
[0089] In an exemplary embodiment, fifth user interface 1500 may initially include third axial slice viewer 1502. When the user selects an artery from third axial slice viewer 1502, the selected blood vessel is presented in third stretched blood vessel image 1504 with vertical bar 1508 denoting the location of the slice presented in third axial slice viewer 1502. Execution of one or more of the processes described with reference to Figs. 2-9 may be performed after selection of the blood vessel to identify the stretched blood vessel presented in third stretched blood vessel image 1504. As a result, using a single "click" the user may trigger a determination of all relevant segments of the blood vessel from the slices and reconstruct the stretched blood vessel for presentation in third stretched blood vessel image 1504.
[0090] Fifth user interface 1500 further may include an axial presentation only button 1510, a new study selection button 1512, a save current screen button 1514, and an exit program button 1516. User selection of axial presentation only button 1510 causes stretched blood vessel image 1504 to be removed from fifth user interface 1500. User selection of new study selection button 1512 causes presentation of a selection window that allows the user to select a new patient study for analysis. User selection of save current screen button 1514 causes presentation of a save window that allows the user to select a location and a name for a file to which the contents of fifth user interface 1500 are saved for review, for printing, for sending with a message, etc. User selection of exit program button 1516 may cause fifth user interface 1500 to close.
[0091] With reference to Fig. 16, exemplary operations associated with assessing a quality of the medical imaging data are described. Additional, fewer, or different operations may be performed, depending on the embodiment. The order of presentation of the operations is not intended to be limiting. The operations described with reference to Fig. 16 may be performed before execution of pathology identification application 112 and/or visualization application 114 (described with reference to Fig. 1) or integrated as part of execution of pathology identification application 112 and/or visualization application 114. For example, the exemplary operations may be performed after operation 210 described with reference to Figs. 2a and 2b to determine a success or failure of the image enhancement process. In an operation 1600, an image resolution is calculated. Image resolution may be stored in the image DICOM header and defined based on the "distance between slices" and "pixel size" fields. In an operation 1602, the calculated image resolution is compared with a quality control parameter such as an image resolution threshold. An exemplary image resolution threshold may be based on an inter-slice distance below 0.7 mm and a pixel size below 0.5 mm. In an operation 1604, a determination is made concerning whether or not the image resolution is high enough. If the calculated image resolution is high enough, processing continues at an operation 1606. If the calculated image resolution is not high enough, processing continues at operation 1624.
[0092] In operation 1606, a determination is made concerning whether or not the heart region is fully visible. A zoom in used in reconstruction should cause the heart region to occupy most of the region of interest (ROI). The heart region may be automatically estimated and, if it is significantly smaller than the ROI, it may be determined that the heart region is not fully visible. If the heart region is fully visible, processing continues at an operation 1608. If the heart region is not fully visible, processing continues at operation 1624. In operation 1608, a determination is made concerning whether or not the contrast injection density is acceptable. The volume of contrast may be estimated by thresholding the image data for example at 100 HU. If the determined volume of contrast is lower than a predefined threshold such as 30 mm x 30 mm x 20 mm, it may be determined that the contrast injection density is unacceptable. If the contrast injection density is acceptable, processing continues at an operation 1610. If the contrast injection density is not acceptable, processing continues at operation 1624.
[0093] The image may be searched for various artifacts using specialized filters. In operation 1610, a determination is made concerning whether or not motion artifacts exist. If motion artifacts do not exist, processing continues at an operation 1612. If motion artifacts exist, processing continues at operation 1624. In operation 1612, a determination is made concerning whether or not reconstruction artifacts exist. If reconstruction artifacts do not exist, processing continues at an operation 1614. If reconstruction artifacts exist, processing continues at operation 1624. In operation 1614, a determination is made concerning whether or not physical artifacts exist. If physical artifacts do not exist, processing continues at an operation 1616. If physical artifacts exist, processing continues at operation 1624.
[0094] In operation 1616, an image signal-to-noise ratio (SNR) is calculated. In an operation 1618, the calculated SNR is compared with a quality control parameter such as a SNR threshold. The image SNR may be computed as [(Mn, -MIB)/σm]2 where Mw is the mean of the intensity in the vessel, MIB is the mean of the intensity of the background, and σIB is the standard deviation of the intensity of the background. The vessel and background locations may be coarsely estimated in order to collect the statistics. In an operation 1620, a determination is made concerning whether or not the calculated SNR is high enough. In an exemplary embodiment, a SNR threshold of 100 dB is used. If the calculated SNR is high enough, processing continues at an operation 1622. In operation 1622, processing continues. In operation 1624, a message is output to the user. For example, a message may be printed from computing device 102, may be displayed at computing device 102, may played audibly from computing device 102, may be transmitted to the user (i.e., e-mail, text message, instant message, voice message, etc.), etc. The message may indicate an assessment of the image data quality based on the test condition which failed. In an operation 1626, processing of the CT image data stops.
[0095] One or more quality control parameters may be selected by the user for example using a user interface presented using pathology identification application 112 and/or visualization application 114. The selected quality control parameters may be defined and/or stored in a configuration file stored at computing device 102 or on another computing device accessible from computing device 102. The quality control parameters may be defined as a function of the type of medical diagnostic equipment used to generate the imaging data. The quality control parameters may be predefined. [0096] Comparison with quality control parameters allows supervision of the quality of the results produced by pathology identification application 112 and/or visualization application 114. A failure of a test comparison performed on a partial result may indicate either a total failure of the automatic processing or a low confidence in the result. In the former case, the processing should be stopped. In the latter case, the confidence level should be propagated to the next processing level and finally presented to the user. Each module of the processing pipeline in addition to the expected output generates a number from zero to one designating the module confidence in the result. For each module, the confidence is computed differently and the computation depends on the particular algorithm. The overall confidence of the system in the final results is computed as the product of confidence levels produced by each module. A failure may be identified when a confidence level exceeds a threshold. The failure in compliance with a required quality level of the input imaging data should prevent the system from automatic processing of the study and warn the operator about the low quality of the acquired data. The operator may decide in such case to perform a repeat examination. In addition, the image quality control system can be used as a separate module (i.e. without pathology identification application 112 and/or visualization application 114) in a medical imaging facility as a tool for automatic acquisition quality control and CT machine calibration.
[0097] The word "exemplary" is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other aspects or designs. Further, for the purposes of this disclosure and unless otherwise specified, "a" or "an" means "one or more".
[0098] The foregoing description of exemplary embodiments of the invention have been presented for purposes of illustration and of description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. The functionality described may be implemented in a single executable or application or may be distributed among modules that differ in number and distribution of functionality from those described herein. Additionally, the order of execution of the functions may be changed depending on the embodiment. The embodiments were chosen and described in order to explain the principles of the invention and as practical applications of the invention to enable one skilled in the art to utilize the invention in various embodiments and with various modifications as suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.

Claims

WHAT IS CLAIMED IS:
1. A system for identifying a data quality issue associated with medical diagnostic data, the system comprising: an imaging apparatus configured to generate imaging data including a heart region and an aorta region of a patient; and a processor operably coupled to the imaging apparatus to receive the generated imaging data, the processor configured to analyze the received imaging data; to identify a data quality issue associated with the received imaging data based on the analysis of the received imaging data; to generate a message including the identified data quality issue; and to send the generated message to a user.
2. Computer-readable instructions that, upon execution by a processor, cause the processor to identify a data quality issue associated with medical diagnostic data, the instructions comprising: receiving imaging data; analyzing the received imaging data; identifying a data quality issue associated with the received imaging data based on the analysis of the received imaging data; generating a message including the identified data quality issue; and sending the generated message to a user.
3. A method of identifying a data quality issue associated with medical diagnostic data, the method comprising: receiving imaging data at a computing device; analyzing the received imaging data at the computing device; identifying a data quality issue associated with the received imaging data based on the analysis of the received imaging data at the computing device; generating a message including the identified data quality issue at the computing device; and sending the generated message to a user.
4. The method of claim 3, wherein sending the generated message to the user is a method selected from the group consisting of printing the generated message, displaying the generated message, and playing audibly the generated message.
5. The method of claim 3, wherein sending the generated message comprises a text message.
6. The method of claim 3, wherein the imaging data is computed tomography data.
7. The method of claim 6, wherein the imaging data is computed tomography angiography data.
8. The method of claim 3, wherein the generated message includes an assessment of the image data quality.
9. The method of claim 8, wherein the assessment is a characteristic selected from the group consisting of a signal-to-noise ratio value, an image resolution value, a physical artifact indicator, a reconstruction artifact indicator, a motion artifact indicator, and a contrast injection density value.
10. The method of claim 8, wherein the assessment of the image data quality is based on a determination that a heart region is not fully visible.
11. The method of claim 3, wherein identifying the data quality issue comprises: comparing the analyzed imaging data with a quality control parameter; and identifying the data quality issue based on the comparison.
12. The method of claim 11 , wherein the quality control parameter is a size of a heart region defined in the analyzed imaging data.
13. The method of claim 11 , wherein the quality control parameter is a location of a heart region in the analyzed imaging data.
14. The method of claim 11 , wherein the quality control parameter is an intensity level of a lung region defined in the analyzed imaging data.
15. The method of claim 11 , wherein the quality control parameter is a volume of contrast defined in the analyzed imaging data.
16. The method of claim 11 , wherein the quality control parameter is a position of an object in a heart region defined from the analyzed imaging data.
17. The method of claim 11 , wherein the quality control parameter is an orientation of an object in a heart region defined from the analyzed imaging data.
18. The method of claim 11 , wherein the quality control parameter is a position of an object relative to an aorta object defined in the analyzed imaging data.
19. The method of claim 11 , wherein the quality control parameter is a size of an object defined in the analyzed imaging data.
20. The method of claim 11 , wherein the quality control parameter is a number of pathologies detected per blood vessel segment defined from the analyzed imaging data.
PCT/IB2007/003193 2006-10-25 2007-10-24 Method and system for automatic quality control used in computerized analysis of ct angiography WO2008050222A2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP07848825A EP2073701A4 (en) 2006-10-25 2007-10-24 Method and system for automatic quality control used in computerized analysis of ct angiography

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US86291206P 2006-10-25 2006-10-25
US60/862,912 2006-10-25
US11/683,767 2007-03-08
US11/683,767 US7940970B2 (en) 2006-10-25 2007-03-08 Method and system for automatic quality control used in computerized analysis of CT angiography

Publications (2)

Publication Number Publication Date
WO2008050222A2 true WO2008050222A2 (en) 2008-05-02
WO2008050222A3 WO2008050222A3 (en) 2009-04-23

Family

ID=39324976

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2007/003193 WO2008050222A2 (en) 2006-10-25 2007-10-24 Method and system for automatic quality control used in computerized analysis of ct angiography

Country Status (3)

Country Link
US (1) US7940970B2 (en)
EP (1) EP2073701A4 (en)
WO (1) WO2008050222A2 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010128412A1 (en) * 2009-05-05 2010-11-11 Koninklijke Philips Electronics, N.V. Automatic assessment of confidence in imaging data
WO2013098074A3 (en) * 2011-12-28 2014-01-23 Siemens Aktiengesellschaft Control method and control system
WO2017136762A1 (en) * 2016-02-03 2017-08-10 Radlogics, Inc. Methods and systems for enhanced measurement of near heart aorta sections
US10062164B2 (en) 2014-04-07 2018-08-28 Mimo Ag Method for the analysis of image data representing a three-dimensional volume of biological tissue

Families Citing this family (34)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8295565B2 (en) * 2007-03-16 2012-10-23 Sti Medical Systems, Llc Method of image quality assessment to produce standardized imaging data
JP5100285B2 (en) * 2007-09-28 2012-12-19 キヤノン株式会社 MEDICAL DIAGNOSIS SUPPORT DEVICE, ITS CONTROL METHOD, PROGRAM, AND STORAGE MEDIUM
WO2009062062A1 (en) * 2007-11-09 2009-05-14 Imacor, Llc Superimposed display of image contours
US20090204338A1 (en) * 2008-02-13 2009-08-13 Nordic Bioscience A/S Method of deriving a quantitative measure of the instability of calcific deposits of a blood vessel
US9064300B2 (en) * 2008-02-15 2015-06-23 Siemens Aktiengesellshaft Method and system for automatic determination of coronory supply regions
JP4671204B2 (en) * 2008-07-09 2011-04-13 ザイオソフト株式会社 Medical image display control apparatus and medical image display control program
DE102009014763B4 (en) 2009-03-25 2018-09-20 Siemens Healthcare Gmbh Method and data processing system for determining the calcium content in coronary vessels
CN105120269B (en) * 2009-08-12 2018-10-19 汤姆森特许公司 For chroma coder in improved frame and decoded method and device
DE102010009701A1 (en) * 2010-03-01 2011-09-01 Siemens Aktiengesellschaft Method for automatically identifying narrow points of coronary blood vessel in heart of patient, involves identifying narrow points in hollow vessel based on automated evaluation of profile course curve of hollow vessel
US9589204B2 (en) * 2010-09-20 2017-03-07 Koninklijke Philips N.V. Quantification of a characteristic of a lumen of a tubular structure
KR20120102447A (en) * 2011-03-08 2012-09-18 삼성전자주식회사 Method and apparatus for diagnostic
EP2518690A1 (en) * 2011-04-28 2012-10-31 Koninklijke Philips Electronics N.V. Medical image processing system and method
EP2727080B1 (en) * 2011-06-29 2019-03-13 The Regents of The University of Michigan Analysis of temporal changes in registered tomographic images
KR20140128940A (en) * 2011-10-10 2014-11-06 트랙터스 코포레이션 Method, apparatus and system for complete examination of tissue with hand-held imaging devices
US9129417B2 (en) * 2012-02-21 2015-09-08 Siemens Aktiengesellschaft Method and system for coronary artery centerline extraction
JP5442799B2 (en) * 2012-04-27 2014-03-12 楽天株式会社 Tag management apparatus, tag management method, tag management program, and computer-readable recording medium storing the program
JP2016506781A (en) * 2013-01-17 2016-03-07 トラクトゥス・コーポレーション Method, apparatus and system for complete examination of tissue with a hand-held imaging device fitted with a camera
US9042613B2 (en) * 2013-03-01 2015-05-26 Heartflow, Inc. Method and system for determining treatments by modifying patient-specific geometrical models
US8824752B1 (en) 2013-03-15 2014-09-02 Heartflow, Inc. Methods and systems for assessing image quality in modeling of patient anatomic or blood flow characteristics
JP2016517746A (en) * 2013-04-30 2016-06-20 トラクトゥス・コーポレーション Method, apparatus and system for complete examination of tissue using a handheld imaging device having position and / or orientation sensors
WO2014210430A1 (en) 2013-06-27 2014-12-31 Tractus Corporation Systems and methods for tissue mapping
US20150164451A1 (en) * 2013-12-18 2015-06-18 Heartflow, Inc. Systems and methods for predicting coronary plaque vulnerability from patient-specific anatomic image data
CN104837007B (en) * 2014-02-11 2018-06-05 阿里巴巴集团控股有限公司 A kind of method and apparatus of digital picture quality classification
US10216762B2 (en) 2014-06-04 2019-02-26 Panasonic Corporation Control method and non-transitory computer-readable recording medium for comparing medical images
US10467464B2 (en) 2016-06-07 2019-11-05 The Neat Company, Inc. Document field detection and parsing
US10650512B2 (en) 2016-06-14 2020-05-12 The Regents Of The University Of Michigan Systems and methods for topographical characterization of medical image data
WO2018029237A1 (en) * 2016-08-10 2018-02-15 Agfa Healthcare Method for automatically detecting systemic arteries in arbitrary field-of-view computed tomography angiography (cta).
DE102016219488A1 (en) 2016-10-07 2018-04-12 Siemens Healthcare Gmbh Method for providing a confidence information
US10861156B2 (en) * 2018-02-28 2020-12-08 Case Western Reserve University Quality control for digital pathology slides
DE102018207632A1 (en) * 2018-05-16 2019-11-21 Siemens Healthcare Gmbh Determining an imaging modality and the parameters therefor
US20220138936A1 (en) * 2019-02-19 2022-05-05 Cedars-Sinai Medical Center Systems and methods for calcium-free computed tomography angiography
DE102019202830A1 (en) * 2019-03-01 2020-09-03 Siemens Healthcare Gmbh Method for evaluating an image quality of a volume image data set
JP7358120B2 (en) * 2019-08-30 2023-10-10 キヤノン株式会社 Information processing device, tomography device, information processing method and program
WO2022164374A1 (en) * 2021-02-01 2022-08-04 Kahraman Ali Teymur Automated measurement of morphometric and geometric parameters of large vessels in computed tomography pulmonary angiography

Family Cites Families (206)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2613509B1 (en) * 1987-04-03 1989-06-09 Thomson Cgr METHOD FOR CALCULATING AND REPRESENTING IMAGES OF VIEWS OF AN OBJECT
US5187658A (en) * 1990-01-17 1993-02-16 General Electric Company System and method for segmenting internal structures contained within the interior region of a solid object
US5457754A (en) 1990-08-02 1995-10-10 University Of Cincinnati Method for automatic contour extraction of a cardiac image
DE69131681T2 (en) 1990-11-22 2000-06-08 Toshiba Kawasaki Kk Computerized diagnostic system for medical use
US5779634A (en) 1991-05-10 1998-07-14 Kabushiki Kaisha Toshiba Medical information processing system for supporting diagnosis
US5343390A (en) 1992-02-28 1994-08-30 Arch Development Corporation Method and system for automated selection of regions of interest and detection of septal lines in digital chest radiographs
EP0616290B1 (en) 1993-03-01 2003-02-05 Kabushiki Kaisha Toshiba Medical information processing system for supporting diagnosis.
US5579767A (en) 1993-06-07 1996-12-03 Prince; Martin R. Method for imaging abdominal aorta and aortic aneurysms
US5435310A (en) 1993-06-23 1995-07-25 University Of Washington Determining cardiac wall thickness and motion by imaging and three-dimensional modeling
FR2708166A1 (en) * 1993-07-22 1995-01-27 Philips Laboratoire Electroniq A method of processing digitized images for the automatic detection of stenoses.
IL106691A (en) 1993-08-13 1998-02-08 Sophis View Tech Ltd System and method for diagnosis of living tissue diseases
US5412563A (en) * 1993-09-16 1995-05-02 General Electric Company Gradient image segmentation method
US6434262B2 (en) 1993-09-29 2002-08-13 Shih-Ping Wang Computer-aided diagnosis system and method
US6574357B2 (en) 1993-09-29 2003-06-03 Shih-Ping Wang Computer-aided diagnosis method and system
US6075879A (en) * 1993-09-29 2000-06-13 R2 Technology, Inc. Method and system for computer-aided lesion detection using information from multiple images
CA2132138C (en) 1993-09-29 2004-01-06 Shih-Ping Wang Computer-aided diagnosis system and method
US5729620A (en) * 1993-09-29 1998-03-17 Wang; Shih-Ping Computer-aided diagnosis system and method
US20020097902A1 (en) 1993-09-29 2002-07-25 Roehrig Jimmy R. Method and system for the display of regions of interest in medical images
US6266435B1 (en) 1993-09-29 2001-07-24 Shih-Ping Wang Computer-aided diagnosis method and system
US5452367A (en) 1993-11-29 1995-09-19 Arch Development Corporation Automated method and system for the segmentation of medical images
US5724968A (en) * 1993-12-29 1998-03-10 First Opinion Corporation Computerized medical diagnostic system including meta function
US5660176A (en) 1993-12-29 1997-08-26 First Opinion Corporation Computerized medical diagnostic and treatment advice system
US5594638A (en) * 1993-12-29 1997-01-14 First Opinion Corporation Computerized medical diagnostic system including re-enter function and sensitivity factors
US6206829B1 (en) * 1996-07-12 2001-03-27 First Opinion Corporation Computerized medical diagnostic and treatment advice system including network access
WO1995020343A1 (en) * 1994-01-28 1995-08-03 Schneider Medical Technologies, Inc. Imaging device and method
US5433199A (en) 1994-02-24 1995-07-18 General Electric Company Cardiac functional analysis method using gradient image segmentation
US5458126A (en) 1994-02-24 1995-10-17 General Electric Company Cardiac functional analysis system employing gradient image segmentation
US5881124A (en) * 1994-03-31 1999-03-09 Arch Development Corporation Automated method and system for the detection of lesions in medical computed tomographic scans
US5734739A (en) * 1994-05-31 1998-03-31 University Of Washington Method for determining the contour of an in vivo organ using multiple image frames of the organ
US5671294A (en) 1994-09-15 1997-09-23 The United States Of America As Represented By The Secretary Of The Navy System and method for incorporating segmentation boundaries into the calculation of fractal dimension features for texture discrimination
WO1996012187A1 (en) 1994-10-13 1996-04-25 Horus Therapeutics, Inc. Computer assisted methods for diagnosing diseases
WO1996016534A2 (en) 1994-11-25 1996-06-06 Sophisview Technologies, Ltd. System and method for diagnosis of living tissue diseases
US5627907A (en) 1994-12-01 1997-05-06 University Of Pittsburgh Computerized detection of masses and microcalcifications in digital mammograms
CA2214101A1 (en) * 1995-03-03 1996-09-12 Ulrich Bick Method and system for the detection of lesions in medical images
US5790690A (en) * 1995-04-25 1998-08-04 Arch Development Corporation Computer-aided method for automated image feature analysis and diagnosis of medical images
JPH10504225A (en) 1995-06-07 1998-04-28 ユニバーシティ オブ フロリダ リサーチ ファウンデーション,インク. An automated method for digital image quantification
US5647360A (en) 1995-06-30 1997-07-15 Siemens Corporate Research, Inc. Digital subtraction angiography for 3D diagnostic imaging
US5889524A (en) * 1995-09-11 1999-03-30 University Of Washington Reconstruction of three-dimensional objects using labeled piecewise smooth subdivision surfaces
US5768413A (en) * 1995-10-04 1998-06-16 Arch Development Corp. Method and apparatus for segmenting images using stochastically deformable contours
US6067372A (en) * 1996-02-22 2000-05-23 University Of Pittsburgh Method and system to enhance robust identification of abnormal regions in radiographs
US6047080A (en) 1996-06-19 2000-04-04 Arch Development Corporation Method and apparatus for three-dimensional reconstruction of coronary vessels from angiographic images
US6909797B2 (en) 1996-07-10 2005-06-21 R2 Technology, Inc. Density nodule detection in 3-D digital images
US7286695B2 (en) 1996-07-10 2007-10-23 R2 Technology, Inc. Density nodule detection in 3-D digital images
US6047090A (en) * 1996-07-31 2000-04-04 U.S. Philips Corporation Method and device for automatic segmentation of a digital image using a plurality of morphological opening operation
US5903664A (en) * 1996-11-01 1999-05-11 General Electric Company Fast segmentation of cardiac images
GB9623573D0 (en) * 1996-11-13 1997-01-08 Philips Electronics Nv Image segmentation
US6687405B1 (en) * 1996-11-13 2004-02-03 Koninklijke Philips Electronics N.V. Image segmentation
US5836877A (en) 1997-02-24 1998-11-17 Lucid Inc System for facilitating pathological examination of a lesion in tissue
US6031374A (en) * 1997-04-11 2000-02-29 Epstein; Frederick H. Method for extracting deformations from velocity-encoded magnetic resonance images of the heart
US6106466A (en) 1997-04-24 2000-08-22 University Of Washington Automated delineation of heart contours from images using reconstruction-based modeling
US6317617B1 (en) 1997-07-25 2001-11-13 Arch Development Corporation Method, computer program product, and system for the automated analysis of lesions in magnetic resonance, mammogram and ultrasound images
US6021404A (en) * 1997-08-18 2000-02-01 Moukheibir; Nabil W. Universal computer assisted diagnosis
US6246784B1 (en) * 1997-08-19 2001-06-12 The United States Of America As Represented By The Department Of Health And Human Services Method for segmenting medical images and detecting surface anomalies in anatomical structures
US6148095A (en) 1997-09-08 2000-11-14 University Of Iowa Research Foundation Apparatus and method for determining three-dimensional representations of tortuous vessels
IT1297396B1 (en) * 1997-12-30 1999-09-01 Francesco Buzzigoli METHOD AND DEVICE FOR THE RECONSTRUCTION OF THREE-DIMENSIONAL IMAGES OF BLOOD VESSELS, IN PARTICULAR OF CORONARY ARTERIES, OR OF OTHER
US6928314B1 (en) 1998-01-23 2005-08-09 Mayo Foundation For Medical Education And Research System for two-dimensional and three-dimensional imaging of tubular structures in the human body
US6317509B1 (en) 1998-02-11 2001-11-13 Analogic Corporation Computed tomography apparatus and method for classifying objects
US6035014A (en) * 1998-02-11 2000-03-07 Analogic Corporation Multiple-stage apparatus and method for detecting objects in computed tomography data
DE19806728A1 (en) * 1998-02-18 1999-08-19 Philips Patentverwaltung Process for two-dimensional imaging of structures for medical diagnostics
US6377832B1 (en) * 1998-03-20 2002-04-23 Georgia Tech Research Corporation System and method for analyzing a medical image
US6385474B1 (en) * 1999-03-19 2002-05-07 Barbara Ann Karmanos Cancer Institute Method and apparatus for high-resolution detection and characterization of medical pathologies
US6396939B1 (en) * 1998-05-28 2002-05-28 Orthosoft Inc. Method and system for segmentation of medical images
US6149595A (en) * 1998-07-02 2000-11-21 Seitz; Walter S. Noninvasive apparatus and method for the determination of cardiac valve function
US6138045A (en) 1998-08-07 2000-10-24 Arch Development Corporation Method and system for the segmentation and classification of lesions
US6708055B2 (en) * 1998-08-25 2004-03-16 University Of Florida Method for automated analysis of apical four-chamber images of the heart
AU6417599A (en) 1998-10-08 2000-04-26 University Of Kentucky Research Foundation, The Methods and apparatus for (in vivo) identification and characterization of vulnerable atherosclerotic plaques
US6771262B2 (en) 1998-11-25 2004-08-03 Siemens Corporate Research, Inc. System and method for volume rendering-based segmentation
US6249590B1 (en) * 1999-02-01 2001-06-19 Eastman Kodak Company Method for automatically locating image pattern in digital images
US6625303B1 (en) 1999-02-01 2003-09-23 Eastman Kodak Company Method for automatically locating an image pattern in digital images using eigenvector analysis
US6320976B1 (en) 1999-04-01 2001-11-20 Siemens Corporate Research, Inc. Computer-assisted diagnosis method and system for automatically determining diagnostic saliency of digital images
US6674894B1 (en) * 1999-04-20 2004-01-06 University Of Utah Research Foundation Method and apparatus for enhancing an image using data optimization and segmentation
US7343195B2 (en) * 1999-05-18 2008-03-11 Mediguide Ltd. Method and apparatus for real time quantitative three-dimensional image reconstruction of a moving organ and intra-body navigation
IL131282A (en) * 1999-08-05 2009-02-11 Orbotech Ltd Apparatus and methods for inspection of objects
US6200268B1 (en) * 1999-09-10 2001-03-13 The Cleveland Clinic Foundation Vascular plaque characterization
US6415046B1 (en) 1999-10-07 2002-07-02 Edmund Kenneth Kerut, Sr. Method and apparatus for the early detection of tissue pathology using wavelet transformation
US7333648B2 (en) 1999-11-19 2008-02-19 General Electric Company Feature quantification from multidimensional image data
AUPQ449899A0 (en) * 1999-12-07 2000-01-06 Commonwealth Scientific And Industrial Research Organisation Knowledge based computer aided diagnosis
US6553356B1 (en) * 1999-12-23 2003-04-22 University Of Pittsburgh - Of The Commonwealth System Of Higher Education Multi-view computer-assisted diagnosis
US6898303B2 (en) * 2000-01-18 2005-05-24 Arch Development Corporation Method, system and computer readable medium for the two-dimensional and three-dimensional detection of lesions in computed tomography scans
US6901156B2 (en) 2000-02-04 2005-05-31 Arch Development Corporation Method, system and computer readable medium for an intelligent search workstation for computer assisted interpretation of medical images
WO2001059707A1 (en) * 2000-02-11 2001-08-16 The Government Of The United States Of America, As Represented By The Secretary, Dept. Of Health And Human Services Vessel delineation in magnetic resonance angiographic images
US6535821B2 (en) * 2000-02-11 2003-03-18 University Of Iowa Research Foundation System and method of bolus-chasing angiography with adaptive real-time computed tomography (CT)
US7583992B2 (en) * 2000-03-30 2009-09-01 Wisconsin Alumni Research Foundation Magnetic resonance angiography with automated vessel segmentation
WO2001078005A2 (en) 2000-04-11 2001-10-18 Cornell Research Foundation, Inc. System and method for three-dimensional image rendering and analysis
ATE396648T1 (en) * 2000-05-09 2008-06-15 Paieon Inc SYSTEM AND METHOD FOR THREE-DIMENTIONAL RECONSTRUCTION OF AN ARTERY
US6408201B1 (en) * 2000-06-09 2002-06-18 General Electric Company Method and apparatus for efficient stenosis identification in peripheral arterial vasculature using MR imaging
US6741880B1 (en) * 2000-06-16 2004-05-25 General Electric Company Method and apparatus for efficient stenosis identification and assessment using MR imaging
US6983063B1 (en) * 2000-06-29 2006-01-03 Siemens Corporate Research, Inc. Computer-aided diagnosis method for aiding diagnosis of three dimensional digital image data
US6980675B2 (en) * 2000-10-18 2005-12-27 Paieon, Inc. Method for processing images of coronary arteries
US6785409B1 (en) 2000-10-24 2004-08-31 Koninklijke Philips Electronics, N.V. Segmentation method and apparatus for medical images using diffusion propagation, pixel classification, and mathematical morphology
US6744911B1 (en) 2000-11-22 2004-06-01 General Electric Company Tomographic segmentation
US6754376B1 (en) 2000-11-22 2004-06-22 General Electric Company Method for automatic segmentation of medical images
US6771803B1 (en) 2000-11-22 2004-08-03 Ge Medical Systems Global Technology Company, Llc Method and apparatus for fitting a smooth boundary to segmentation masks
WO2002056240A1 (en) 2000-11-22 2002-07-18 R2 Technology, Inc. Graphical user interface for display of anatomical information
US6643533B2 (en) 2000-11-28 2003-11-04 Ge Medical Systems Global Technology Company, Llc Method and apparatus for displaying images of tubular structures
US7072515B2 (en) 2001-01-26 2006-07-04 Rensselaer Polytechnic Institute Method and apparatus for automatically tracing line-structure images
CA2437883C (en) * 2001-03-02 2009-05-19 Yoram Palti Method and apparatus for detecting arterial stenosis
DE10297689B4 (en) * 2001-05-01 2007-10-18 The General Hospital Corp., Boston Method and device for the determination of atherosclerotic coating by measurement of optical tissue properties
JP2004520923A (en) 2001-06-20 2004-07-15 コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ How to segment digital images
US6845260B2 (en) 2001-07-18 2005-01-18 Koninklijke Philips Electronics N.V. Automatic vessel indentification for angiographic screening
JP2003038476A (en) * 2001-08-03 2003-02-12 Konica Corp Computer-aided diagnostic system
US6795521B2 (en) 2001-08-17 2004-09-21 Deus Technologies Llc Computer-aided diagnosis system for thoracic computer tomography images
FR2829268A1 (en) * 2001-09-04 2003-03-07 Koninkl Philips Electronics Nv IMAGE PROCESSING METHOD FOR DIGITIZED SUBTRACTIVE ANGIOGRAPHY
JP4319031B2 (en) * 2001-09-06 2009-08-26 コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ Object segmentation method and apparatus
JP3704492B2 (en) 2001-09-11 2005-10-12 テラリコン・インコーポレイテッド Reporting system in network environment
DK2793189T3 (en) 2001-10-03 2017-02-06 Retinalyze As Assessment of lesions in an image
US6985612B2 (en) * 2001-10-05 2006-01-10 Mevis - Centrum Fur Medizinische Diagnosesysteme Und Visualisierung Gmbh Computer system and a method for segmentation of a digital image
US7158692B2 (en) 2001-10-15 2007-01-02 Insightful Corporation System and method for mining quantitive information from medical images
WO2003034176A2 (en) 2001-10-16 2003-04-24 The University Of Chicago Computer-aided detection of three-dimensional lesions
US6947040B2 (en) 2001-10-23 2005-09-20 Siemens Corporate Research, Inc. Vessel detection by mean shift based ray propagation
US6701170B2 (en) * 2001-11-02 2004-03-02 Nellcor Puritan Bennett Incorporated Blind source separation of pulse oximetry signals
US6842638B1 (en) * 2001-11-13 2005-01-11 Koninklijke Philips Electronics N.V. Angiography method and apparatus
US7127096B2 (en) * 2001-11-20 2006-10-24 Accuimage Diagnostics Corp. Method and software for improving coronary calcium scoring consistency
US7012603B2 (en) 2001-11-21 2006-03-14 Viatronix Incorporated Motion artifact detection and correction
CA2466809A1 (en) 2001-11-21 2003-06-05 Viatronix Incorporated System and method for visualization and navigation of three-dimensional medical images
US6728566B1 (en) * 2001-11-21 2004-04-27 Koninklijke Philips Electronics, N.V. Vessel tracking and tree extraction method and apparatus
US6782284B1 (en) 2001-11-21 2004-08-24 Koninklijke Philips Electronics, N.V. Method and apparatus for semi-automatic aneurysm measurement and stent planning using volume image data
US7336809B2 (en) * 2001-11-23 2008-02-26 R2 Technology, Inc. Segmentation in medical images
US7386153B2 (en) 2001-11-23 2008-06-10 Infinitt Co., Ltd. Medical image segmentation apparatus and method thereof
US8649843B2 (en) 2001-11-24 2014-02-11 Ben A. Arnold Automated calcium scoring of the aorta
WO2003046833A2 (en) 2001-11-24 2003-06-05 Image Analysis, Inc. Automatic detection and quantification of coronary and aortic calcium
US6978039B2 (en) 2001-12-13 2005-12-20 General Electric Company Method and system for segmentation of medical images
WO2003079137A2 (en) 2001-12-28 2003-09-25 Modrovich Ivan Endre Diagnostic information systems
US7136516B2 (en) 2002-01-25 2006-11-14 General Electric Company Method and system for segmenting magnetic resonance images
AU2003216295A1 (en) 2002-02-15 2003-09-09 The Regents Of The University Of Michigan Lung nodule detection and classification
GB0205000D0 (en) 2002-03-04 2002-04-17 Isis Innovation Unsupervised data segmentation
US7046835B2 (en) 2002-03-07 2006-05-16 Ge Medical Systems Global Technology Company Llc Method and system for processing vascular radiographic images which have been reconstructed by three-dimensional modelling
US20030190063A1 (en) 2002-03-08 2003-10-09 Acharya Kishore C. Method and system for performing coronary artery calcification scoring
CN100405973C (en) 2002-03-14 2008-07-30 Netkisr有限公司 System and method for analyzing and displaying computed tomography data
US7142208B2 (en) 2002-03-23 2006-11-28 Koninklijke Philips Electronics, N.V. Method for interactive segmentation of a structure contained in an object
US6819790B2 (en) 2002-04-12 2004-11-16 The University Of Chicago Massive training artificial neural network (MTANN) for detecting abnormalities in medical images
US7263214B2 (en) 2002-05-15 2007-08-28 Ge Medical Systems Global Technology Company Llc Computer aided diagnosis from multiple energy images
US7295691B2 (en) 2002-05-15 2007-11-13 Ge Medical Systems Global Technology Company, Llc Computer aided diagnosis of an image set
JP4138371B2 (en) 2002-06-06 2008-08-27 富士フイルム株式会社 Anatomical feature position detecting device, recording medium, subject structure measuring device, and recording medium
DE10229113A1 (en) 2002-06-28 2004-01-22 Siemens Ag Process for gray value-based image filtering in computer tomography
US20050033139A1 (en) * 2002-07-09 2005-02-10 Deus Technologies, Llc Adaptive segmentation of anatomic regions in medical images with fuzzy clustering
US7418123B2 (en) * 2002-07-12 2008-08-26 University Of Chicago Automated method and system for computerized image analysis for prognosis
US6922462B2 (en) 2002-07-31 2005-07-26 Ge Medical Systems Global Technology Company, Llc Method, system and computer product for plaque characterization
US20040133100A1 (en) 2002-08-23 2004-07-08 Morteza Naghavi Novel risk assessment method based upon coronary calcification distribution pattern imaged by computed tomography
AU2003270654A1 (en) 2002-09-12 2004-04-30 Baylor College Of Medecine System and method for image segmentation
US20040068167A1 (en) * 2002-09-13 2004-04-08 Jiang Hsieh Computer aided processing of medical images
US6574304B1 (en) * 2002-09-13 2003-06-03 Ge Medical Systems Global Technology Company, Llc Computer aided acquisition of medical images
US7668370B2 (en) 2002-09-19 2010-02-23 Koninklijke Philips Electronics N.V. Segmenting a series of 2D or 3D images
WO2004036500A2 (en) * 2002-10-16 2004-04-29 Koninklijke Philips Electronics N.V. Hierarchical image segmentation
DE10249643A1 (en) 2002-10-24 2004-05-13 Siemens Ag Coronary heart disease diagnosis and treatment method in which plaque deposition in blood vessels of interest is monitored over time and compared with reference values stored in the memory of a data processing unit
US7244230B2 (en) 2002-11-08 2007-07-17 Siemens Medical Solutions Usa, Inc. Computer aided diagnostic assistance for medical imaging
EP1420367A1 (en) * 2002-11-15 2004-05-19 MeVis GmbH A method for coloring of voxels and image data processing and visualization system
US7123760B2 (en) * 2002-11-21 2006-10-17 General Electric Company Method and apparatus for removing obstructing structures in CT imaging
US7177453B2 (en) * 2002-11-26 2007-02-13 General Electric Company Method and apparatus for partitioning a volume
US6597762B1 (en) 2002-11-27 2003-07-22 Ge Medical Systems Global Technology Co., Llc Method and apparatus of lesion detection and validation based on multiple reviews of a CT image
US7298883B2 (en) 2002-11-29 2007-11-20 University Of Chicago Automated method and system for advanced non-parametric classification of medical images and lesions
US7221787B2 (en) 2002-12-10 2007-05-22 Eastman Kodak Company Method for automated analysis of digital chest radiographs
US7239730B2 (en) 2003-01-29 2007-07-03 Ge Medical Systems Global Technology Company, Llc Method and apparatus for volume scoring calcification concentrations of a CT scan
US6937776B2 (en) 2003-01-31 2005-08-30 University Of Chicago Method, system, and computer program product for computer-aided detection of nodules with three dimensional shape enhancement filters
US6754380B1 (en) 2003-02-14 2004-06-22 The University Of Chicago Method of training massive training artificial neural networks (MTANN) for the detection of abnormalities in medical images
US6950544B2 (en) 2003-02-14 2005-09-27 Virtualscopics, Llc Automated measurement of anatomical structures in medical imaging
US7333644B2 (en) 2003-03-11 2008-02-19 Siemens Medical Solutions Usa, Inc. Systems and methods for providing automatic 3D lesion segmentation and measurements
AU2004251360A1 (en) * 2003-06-25 2005-01-06 Siemens Medical Solutions Usa, Inc. Automated regional myocardial assessment for cardiac imaging
US7912528B2 (en) * 2003-06-25 2011-03-22 Siemens Medical Solutions Usa, Inc. Systems and methods for automated diagnosis and decision support for heart related diseases and conditions
EP1644869A4 (en) 2003-07-15 2011-01-26 Medical Metrx Solutions Inc Using statistical process control (spc) to demonstrate solution convergence in a technician guided segmentation system
CA2535133C (en) * 2003-08-13 2011-03-08 Siemens Medical Solutions Usa, Inc. Computer-aided decision support systems and methods
US7657299B2 (en) * 2003-08-21 2010-02-02 Ischem Corporation Automated methods and systems for vascular plaque detection and analysis
US20050136549A1 (en) 2003-10-30 2005-06-23 Bioimagene, Inc. Method and system for automatically determining diagnostic saliency of digital images
US7346201B2 (en) * 2003-09-26 2008-03-18 Virtualscopics Llc Semi-automated measurement of anatomical structures using statistical and morphological priors
US7251374B2 (en) * 2003-10-03 2007-07-31 Confirma, Inc. System and method for hierarchical analysis of contrast enhanced medical imaging information
US20050074150A1 (en) * 2003-10-03 2005-04-07 Andrew Bruss Systems and methods for emulating an angiogram using three-dimensional image data
US7233329B2 (en) * 2003-11-03 2007-06-19 Siemens Corporate Research, Inc. Rendering for coronary visualization
US7649176B2 (en) * 2003-11-14 2010-01-19 Siemens Medical Solutions Usa, Inc. Method for improving clinical data quality in positron emission tomography
JP4118786B2 (en) * 2003-11-14 2008-07-16 ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー Imaging support system
US7574247B2 (en) 2003-11-17 2009-08-11 Siemens Medical Solutions Usa, Inc. Automatic coronary isolation using a n-MIP ray casting technique
FR2862786B1 (en) * 2003-11-21 2006-01-06 Ge Med Sys Global Tech Co Llc METHOD FOR ELASTICALLY REPRESENTING AN IMAGE APPLIED TO DIGITALIZED SUBTRACT ANGIOGRAPHY
US7676257B2 (en) * 2003-11-25 2010-03-09 General Electric Company Method and apparatus for segmenting structure in CT angiography
US7447341B2 (en) * 2003-11-26 2008-11-04 Ge Medical Systems Global Technology Company, Llc Methods and systems for computer aided targeting
US20050110791A1 (en) * 2003-11-26 2005-05-26 Prabhu Krishnamoorthy Systems and methods for segmenting and displaying tubular vessels in volumetric imaging data
WO2005055148A1 (en) 2003-11-29 2005-06-16 Vital Images, Inc. Segmented volume rendering using a programmable graphics pipeline
US20050256400A1 (en) 2003-12-03 2005-11-17 Bhargav Raman Method to identify arterial and venous vessels
US7330576B2 (en) 2003-12-03 2008-02-12 The Board Of Trustees Of The Leland Stanford Junior University Quantification method of vessel calcification
US7189000B2 (en) * 2003-12-22 2007-03-13 Kabushiki Kaisha Toshiba Image-quality control system
US7444011B2 (en) * 2004-02-10 2008-10-28 University Of Chicago Imaging system performing substantially exact reconstruction and using non-traditional trajectories
WO2005079306A2 (en) * 2004-02-13 2005-09-01 University Of Chicago Method, system, and computer software product for feature-based correlation of lesions from multiple images
US7526115B2 (en) 2004-02-23 2009-04-28 Siemens Medical Solutions Usa, Inc. System and method for toboggan based object segmentation using divergent gradient field response in images
US7567696B2 (en) 2004-03-02 2009-07-28 Siemens Medical Solutions Usa, Inc. System and method for detecting the aortic valve using a model-based segmentation technique
JP2005253685A (en) 2004-03-11 2005-09-22 Konica Minolta Medical & Graphic Inc Diagnostic imaging support device and program
US7327880B2 (en) 2004-03-12 2008-02-05 Siemens Medical Solutions Usa, Inc. Local watershed operators for image segmentation
US20050201606A1 (en) 2004-03-12 2005-09-15 Kazunori Okada 3D segmentation of targets in multislice image
CN101420905B (en) 2004-04-14 2013-01-02 美国医软科技公司 Lesion marking and characterization quality assurance method and system
CN101421745B (en) 2004-04-15 2016-05-11 美国医软科技公司 Spatial-temporal lesion detects, and cuts apart and diagnostic information extraction system and method
US7447344B2 (en) 2004-04-16 2008-11-04 Siemens Medical Solutions Usa, Inc. System and method for visualization of pulmonary emboli from high-resolution computed tomography images
FR2869438B1 (en) 2004-04-21 2006-10-13 Ge Med Sys Global Tech Co Llc METHOD FOR AUTOMATIC SEGMENTATION OF CARDIAC CAVITIES
US7324661B2 (en) 2004-04-30 2008-01-29 Colgate-Palmolive Company Computer-implemented system and method for automated and highly accurate plaque analysis, reporting, and visualization
US7620501B2 (en) 2004-05-06 2009-11-17 Siemens Medical Solutions Usa, Inc. Artery-vein separation and vessel modeling
US7397935B2 (en) 2004-05-10 2008-07-08 Mediguide Ltd. Method for segmentation of IVUS image sequences
US20050254546A1 (en) 2004-05-12 2005-11-17 General Electric Company System and method for segmenting crowded environments into individual objects
US20060009694A1 (en) * 2004-05-17 2006-01-12 Yousefzadeh David K Methods of attenuating internal radiation exposure
GB2414357A (en) 2004-05-18 2005-11-23 Medicsight Plc Nodule boundary detection
US20050259854A1 (en) 2004-05-21 2005-11-24 University Of Chicago Method for detection of abnormalities in three-dimensional imaging data
JP4179510B2 (en) 2004-05-31 2008-11-12 株式会社産学連携機構九州 Inspection support device and inspection support program
DE102004027710A1 (en) 2004-06-07 2006-01-26 Siemens Ag Method for the automatic detection of a structure in medical imaging, computer tomography device, workstation and computer program product
GB2461199B (en) 2004-06-23 2010-04-28 Medicsight Plc Lesion extent determination in a CT scan image
US20060036167A1 (en) * 2004-07-03 2006-02-16 Shina Systems Ltd. Vascular image processing
GB2416223A (en) * 2004-07-15 2006-01-18 Medicsight Plc Quantification of coronary artery calcification
US20060018524A1 (en) * 2004-07-15 2006-01-26 Uc Tech Computerized scheme for distinction between benign and malignant nodules in thoracic low-dose CT
DE102004036726A1 (en) 2004-07-29 2006-03-16 Siemens Ag Method and device for visualizing deposits in blood vessels, in particular in coronary vessels
US7555152B2 (en) 2005-01-06 2009-06-30 Siemens Medical Solutions Usa, Inc. System and method for detecting ground glass nodules in medical images
US7532214B2 (en) * 2005-05-25 2009-05-12 Spectra Ab Automated medical image visualization using volume rendering with local histograms
US20080008366A1 (en) * 2006-06-20 2008-01-10 Vladimir Desh Simultaneous visualization, analysis and navigation of multi-modality medical imaging data

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of EP2073701A4 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010128412A1 (en) * 2009-05-05 2010-11-11 Koninklijke Philips Electronics, N.V. Automatic assessment of confidence in imaging data
US9317911B2 (en) 2009-05-05 2016-04-19 Koninklijke Philips N.V. Automatic assessment of confidence in imaging data
WO2013098074A3 (en) * 2011-12-28 2014-01-23 Siemens Aktiengesellschaft Control method and control system
CN104125841A (en) * 2011-12-28 2014-10-29 西门子公司 Control method and control system
KR101774095B1 (en) * 2011-12-28 2017-09-01 지멘스 악티엔게젤샤프트 Control method and control system
US9760830B2 (en) 2011-12-28 2017-09-12 Siemens Aktiengesellschaft Control method and control system
US10062164B2 (en) 2014-04-07 2018-08-28 Mimo Ag Method for the analysis of image data representing a three-dimensional volume of biological tissue
WO2017136762A1 (en) * 2016-02-03 2017-08-10 Radlogics, Inc. Methods and systems for enhanced measurement of near heart aorta sections

Also Published As

Publication number Publication date
EP2073701A4 (en) 2011-10-05
WO2008050222A3 (en) 2009-04-23
EP2073701A2 (en) 2009-07-01
US20080219530A1 (en) 2008-09-11
US7940970B2 (en) 2011-05-10

Similar Documents

Publication Publication Date Title
US7940970B2 (en) Method and system for automatic quality control used in computerized analysis of CT angiography
US7940977B2 (en) Method and system for automatic analysis of blood vessel structures to identify calcium or soft plaque pathologies
US7860283B2 (en) Method and system for the presentation of blood vessel structures and identified pathologies
US8103074B2 (en) Identifying aorta exit points from imaging data
US7873194B2 (en) Method and system for automatic analysis of blood vessel structures and pathologies in support of a triple rule-out procedure
JP4728627B2 (en) Method and apparatus for segmenting structures in CT angiography
US7567696B2 (en) System and method for detecting the aortic valve using a model-based segmentation technique
Aykac et al. Segmentation and analysis of the human airway tree from three-dimensional X-ray CT images
Sluimer et al. Computer analysis of computed tomography scans of the lung: a survey
NL1034718C2 (en) Method and system for creating and using an influence atlas.
EP2279490B1 (en) Automatic detection and accurate segmentation of abdominal aortic aneurysm
JP5081390B2 (en) Method and system for monitoring tumor burden
US20090012382A1 (en) Method and system for detection of obstructions in vasculature
JP2022070934A (en) Stent detecting method
US20050240094A1 (en) System and method for visualization of pulmonary emboli from high-resolution computed tomography images
US20100166283A1 (en) Method for determining a 2d contour of a vessel structure imaged in 3d image data
CN108876794A (en) Aneurysm in volumetric image data with carry being isolated for tumor blood vessel
CN101563708B (en) For the apparatus and method for the false positive for indicating to be computed machine testing in medical imaging data
WO2021117043A1 (en) Automatic stenosis detection
EP2074551A2 (en) Method and system for automatic analysis of blood vessel structures and pathologies
Larralde et al. Evaluation of a 3D segmentation software for the coronary characterization in multi-slice computed tomography
CN114469153B (en) Angiography device and equipment based on CT (computed tomography) image and computer readable medium
Lavi et al. Single-seeded coronary artery tracking in CT angiography
US20220039769A1 (en) Method, device, and system for determining abnormality in myocardium region
EP3244798B1 (en) Adaptive segmentation for rotational c-arm computed tomography with a reduced angular range

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 07848825

Country of ref document: EP

Kind code of ref document: A2

WWE Wipo information: entry into national phase

Ref document number: 2007848825

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE