WO2014112338A1 - 医用画像処理装置および方法並びにプログラム - Google Patents

医用画像処理装置および方法並びにプログラム Download PDF

Info

Publication number
WO2014112338A1
WO2014112338A1 PCT/JP2014/000076 JP2014000076W WO2014112338A1 WO 2014112338 A1 WO2014112338 A1 WO 2014112338A1 JP 2014000076 W JP2014000076 W JP 2014000076W WO 2014112338 A1 WO2014112338 A1 WO 2014112338A1
Authority
WO
WIPO (PCT)
Prior art keywords
region
medical image
image data
extraction
left ventricular
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/JP2014/000076
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
赤堀　貞登
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fujifilm Corp
Original Assignee
Fujifilm Corp
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 Fujifilm Corp filed Critical Fujifilm Corp
Publication of WO2014112338A1 publication Critical patent/WO2014112338A1/ja
Priority to US14/800,704 priority Critical patent/US10019804B2/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0033—Features or image-related aspects of imaging apparatus, e.g. for MRI, optical tomography or impedance tomography apparatus; Arrangements of imaging apparatus in a room
    • A61B5/0035—Features or image-related aspects of imaging apparatus, e.g. for MRI, optical tomography or impedance tomography apparatus; Arrangements of imaging apparatus in a room adapted for acquisition of images from more than one imaging mode, e.g. combining MRI and optical tomography
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0033—Features or image-related aspects of imaging apparatus, e.g. for MRI, optical tomography or impedance tomography apparatus; Arrangements of imaging apparatus in a room
    • A61B5/004—Features or image-related aspects of imaging apparatus, e.g. for MRI, optical tomography or impedance tomography apparatus; Arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part
    • A61B5/0044—Features or image-related aspects of imaging apparatus, e.g. for MRI, optical tomography or impedance tomography apparatus; Arrangements of imaging apparatus in a room adapted for image acquisition of a particular organ or body part for the heart
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/149—Segmentation; Edge detection involving deformable models, e.g. active contour models
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/162—Segmentation; Edge detection involving graph-based methods
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/174—Segmentation; Edge detection involving the use of two or more images
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
    • A61B6/503—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications for diagnosis of the heart
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/04—Indexing scheme for image data processing or generation, in general involving 3D image data
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20072—Graph-based image processing
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20092—Interactive image processing based on input by user
    • G06T2207/20101—Interactive definition of point of interest, landmark or seed
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20128—Atlas-based segmentation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30048—Heart; Cardiac

Definitions

  • the present invention relates to a medical image processing apparatus, method, and program for extracting a left ventricular region, a right ventricular region, and the like from medical image data including a heart image.
  • Non-patent Documents 1 to 3 Conventionally, various methods for extracting a region such as the left ventricle of a heart from medical images such as CT (Computed Tomography), MRI (Magnetic Resonance Imaging), and ultrasound images have been proposed (Non-patent Documents 1 to 3). reference).
  • cardiac functions such as ejection fraction, end-diastolic volume, end-systolic volume, stroke volume, cardiac output, and myocardial weight are further analyzed. be able to.
  • Non-Patent Document 1 and Non-Patent Document 2 propose a method of simultaneously extracting a plurality of rooms in the heart.
  • Non-Patent Document 3 by extracting each room individually, variations are made smaller than when each room is extracted simultaneously from the entire heart.
  • an object of the present invention is to provide a medical image processing apparatus, method, and program capable of extracting each region corresponding to each room of the heart with high accuracy.
  • the medical image processing apparatus of the present invention extracts a medical image data acquisition unit that acquires medical image data including a heart image, and a left ventricular region of the heart in the medical image data acquired by the medical image data acquisition unit, A region extraction processing unit that performs region extraction processing for extracting at least one of the right ventricular region, the left atrial region, and the right atrial region of the heart based on the extraction result of the left ventricular region; To do.
  • the region extraction processing unit can limit the pixels of the medical image data to be subjected to region extraction processing based on the extraction result of the left ventricular region.
  • the region extraction processing unit can perform region extraction processing by setting an evaluation function based on the extraction result of the left ventricular region and obtaining an optimal solution of the evaluation function.
  • the region extraction processing unit can extract the right atrial region or the left atrial region after extracting the right ventricular region based on the extraction result of the left ventricular region.
  • the medical image data acquisition unit acquires a plurality of medical image data having different phases in one cardiac cycle of a beating heart
  • the region extraction processing unit sets one of the plurality of medical image data in the expansion period.
  • a left ventricular region in one piece of medical image data is extracted, and at least one of a right ventricular region, a left atrial region, and a right atrial region in one piece of medical image data in a diastole is extracted based on the extraction result of the left ventricular region.
  • One region is extracted, and based on the extraction result of one piece of medical image data in the expansion period, at least one region in the medical image data having a phase other than one piece of medical image data in the expansion period is extracted. it can.
  • the region extraction processing unit extracts a region of the left ventricular region in one piece of medical image data at the end diastole from the plurality of medical image data, and based on the extraction result of the left ventricular region, At least one region of the right ventricular region, the left atrial region, and the right atrial region in one piece of medical image data is extracted, and 1 in the end diastole is based on the extraction result of one piece of medical image data in the end diastole.
  • the at least one region in the medical image data having a phase other than the sheet of medical image data can be extracted.
  • the region extraction processing unit accepts manual correction of the contour point of the region, and sets the corrected contour point as the corrected contour point. Based on this, it is possible to perform again the region extraction process of the medical image data having a phase other than the one medical image data in the expansion period.
  • the region extraction processing unit can detect a plurality of landmarks on the medical image data and extract a left ventricular region based on the detected result of the detected landmarks.
  • the region extraction processing unit may detect that the left ventricle region or the at least one region cannot be extracted or the extracted left ventricular region or the at least one region is abnormal based on the detection results of the plurality of landmarks. In some cases, manual correction of the landmark is accepted, and the region extraction process can be performed again based on the corrected landmark.
  • the region extraction processing unit accepts manual correction of the contour point of the left ventricular region, and based on the corrected contour point, the region of the left ventricular region The extraction process can be performed again.
  • three-dimensional image data can be used as medical image data.
  • the medical image processing method of the present invention acquires medical image data for acquiring medical image data including a heart image, extracts a left ventricular region of the heart in the acquired medical image data, and extracts the left ventricular region. Based on the above, at least one of the right ventricular region, the left atrial region, and the right atrial region of the heart is extracted.
  • the medical image processing program of the present invention uses a computer to extract a medical image data acquisition unit that acquires medical image data including a heart image and a left ventricular region of the heart in the medical image data acquired by the medical image data acquisition unit. And, based on the extraction result of the left ventricle region, to function as a region extraction processing unit that performs region extraction processing for extracting at least one of the right ventricular region, the left atrial region, and the right atrial region of the heart.
  • medical image data including a heart image is acquired, the left ventricular region of the heart in the acquired medical image data is extracted, and the extraction result of the left ventricular region Since at least one of the right ventricular region, left atrial region, and right atrial region of the heart is extracted based on the above, each region of the heart can be extracted with high accuracy.
  • the shape of the left ventricle region of the heart is relatively simple, it can be extracted with high accuracy by extracting it alone. Then, by extracting the other region using the extraction result of the left ventricle region as reliable information, the extraction accuracy of the other region can be improved, and the extraction error can be reduced.
  • FIG. 1 is a block diagram showing a schematic configuration of a medical image diagnosis support system using an embodiment of a medical image processing apparatus and method and a program according to the present invention.
  • FIG. 1 Flowchart for explaining the operation of a medical image diagnosis support system using an embodiment of the medical image processing apparatus and method and program of the present invention.
  • the figure which shows an example of medical image data containing the image of a heart The figure which shows the left ventricle area
  • the figure which shows an example of the several landmark M on medical image data The figure which shows an example of the several landmark M on medical image data Diagram for explaining region extraction processing using the graph cut method
  • region extraction process using an atlas image The figure for demonstrating the method to perform the area
  • FIG. 1 is a block diagram showing a schematic configuration of a medical image diagnosis support system using this embodiment.
  • the medical image diagnosis support system 1 of the present embodiment includes a medical image processing device 10, a display 20, an input device 30, and a medical image data storage server 40, as shown in FIG.
  • the medical image processing apparatus 10 is configured by installing the medical image processing program of the present embodiment in a computer.
  • the medical image processing apparatus 10 includes a central processing unit (CPU) and a semiconductor memory, a hard disk in which the medical image processing program of this embodiment is installed, a storage device such as an SSD (Solid State Drive), and the like.
  • the medical image data acquisition unit 11, the region extraction processing unit 12, and the display control unit 13 as illustrated in FIG. 1 are configured by hardware. Then, the respective units operate by the medical image processing program installed in the hard disk being executed by the central processing unit.
  • the medical image data acquisition unit 11 acquires medical image data including a heart image taken in advance.
  • medical image data for example, tomographic image data output from a CT apparatus, an MRI apparatus, an MS (Multi-Slice) CT apparatus, a cone beam CT apparatus, an ultrasonic imaging apparatus, or the like, reconstructed from this tomographic image data
  • volume data There is volume data.
  • the medical image data is stored in advance together with the identification information of the subject in the medical image data storage server 40, and the medical image data acquisition unit 11 uses the medical information corresponding to the identification information of the subject input through the input device 30.
  • the image data is read from the medical image data storage server 40.
  • the region extraction processing unit 12 performs region extraction processing using the graph cut method on the medical image data acquired by the medical image data acquisition unit 11. Then, the region extraction processing unit 12 performs region extraction processing to first extract the left ventricular region of the heart in the medical image data, and then, based on the extraction result of the left ventricular region, the right ventricular region of the heart Extracting at least one region of the left atrial region and the right atrial region. In the present embodiment, the left ventricle region is extracted, then the right ventricle region is extracted, and then the left atrial region and the right atrial region are extracted in this order. A specific method for extracting these areas will be described in detail later.
  • the display control unit 13 displays on the display 20 the tomographic image data acquired by the medical image data acquisition unit 11 and the voxel model or surface model obtained by performing volume rendering or surface rendering on the volume data.
  • the display control unit 13 receives the extraction result of each region in the region extraction processing unit 12, and displays an image in which each region is color-coded on the display 20 based on the extraction result.
  • the color-coded images are displayed superimposed on the tomographic image, voxel model, or surface model.
  • the input device 30 receives a user's input of predetermined information, and is constituted by a pointing device such as a keyboard or a mouse.
  • the identification information of the subject is input in the input device 30, and the medical image data acquisition unit 11 of the medical image processing apparatus 10 converts the medical image data corresponding to the input identification information of the subject to the medical image data. It is read from the storage server 40 and acquired (S10). Here, it is assumed that medical image data captured by the CT apparatus is read out and acquired.
  • FIG. 3 shows an example of medical image data including a heart image acquired by the medical image data acquisition unit 11.
  • FIG. 4 shows the left ventricular region LV of the heart in the medical image data shown in FIG.
  • a plurality of landmarks M on the medical image data are designated and set by the user using the input device 30.
  • the number of landmarks M it is desirable to set and input at least three points as shown in FIG.
  • the position of the landmark M a position of a mitral valve or an aortic valve existing in the left ventricular region of the heart, a position of the apex of the left ventricular region, or the like is designated and inputted.
  • the user sets and inputs the landmark M.
  • the present invention is not limited to this, and the mitral valve, the aortic valve, the apex position of the left ventricular region, and the like are automatically detected using pattern matching. You may make it extract to.
  • the landmark M is also used when extracting the right ventricular region, left atrial region, and right atrial region of the heart. Assume that five landmarks M as shown in FIG. For the remaining two points, for example, the position of the tricuspid valve and the position of the heart base near the entrance of the aorta and superior vena cava are set automatically or manually.
  • the region extraction processing unit 12 sets an initial region of the left ventricular region based on the position information of the landmark M set and input by the user or automatically extracted.
  • the initial region of the left ventricle region is a region in which the left ventricle region is roughly extracted.
  • various known methods can be used.
  • the left ventricular region has a conical shape, and has a simple shape as compared with other right ventricular regions and left and right atrial regions, and therefore can be extracted with high accuracy.
  • the method described in Non-Patent Document 3 described above may be used.
  • the initial region of the left ventricle region may be set using atlas image data including a heart image.
  • the atlas image data is created by a doctor using a general anatomical chart, for example, and represents a typical heart shape.
  • the region extraction processing unit 12 uses the landmark set and input automatically or manually on the atlas image data and the landmark M shown in FIG. 6 set and input on the medical image data to be extracted. Used to perform rough alignment between these image data.
  • the landmark on the atlas image data is set automatically or manually as a landmark corresponding to the landmark M shown in FIG.
  • non-rigid alignment is performed to obtain the correspondence between the pixels of the two image data as a matching function.
  • a region on the medical image data corresponding to the left ventricular region which is the correct region on the atlas image data is obtained, and this region is set as the initial region of the left ventricular region.
  • the region extraction processing unit 12 generates a graphical model in the graph cut method using the pixels in the initial region of the left ventricular region extracted as described above and the pixels around the initial region as extraction target pixels. To do. Then, the final left ventricular region is extracted by performing region extraction processing using the graph cut method.
  • region extraction processing using the graph cut method will be described.
  • the region extraction processing unit 12 includes a node N ijk that represents each pixel to be processed as described above, and a node S that represents a label that each pixel can take (here, the left ventricular region and other regions), T, which is a link that connects T, n-link that is a link that connects nodes of adjacent pixels, and a node N ijk that represents each pixel and a node S that represents a left ventricular region or a node T that represents a region other than the left ventricular region.
  • a graphical model composed of -link is generated (see the left figure in FIG. 7; however, FIG. 7 shows the division of a two-dimensional area for easy understanding).
  • n-link represents the probability that adjacent pixels are voxels in the same region, and the probability can be obtained based on, for example, a difference in pixel values between the adjacent pixels.
  • the t-link connecting the node N ijk representing each pixel and the node T representing a region other than the left ventricular region represents the probability that the pixel is a pixel included in the surrounding region other than the left ventricular region.
  • the t-link connecting the node N ijk representing each pixel and the node S representing the left ventricular region represents the probability that the pixel is a pixel included in the left ventricular region.
  • n-link and t-link can be expressed as a cost function representing the likelihood.
  • region are extracted separately as a left ventricle area
  • the t-link connecting the node N ijk and the node T representing the area other than the contrast area is, for example, a CT value around the contrast area where the pixel value is statistically acquired in advance. It can be calculated based on the determination result of whether or not it is within the range of values.
  • the t-link connecting the node N ijk and the node S representing the contrast area is based on, for example, a determination result of whether or not the pixel value is within the CT value range of the contrast area statistically acquired in advance. Can be calculated.
  • one of two t-links connecting a node representing each pixel and a node S representing a contrast region or a node T representing a region other than the contrast region is cut, and adjacent pixels having different labels are connected to each other.
  • the n-link to be connected it is divided into a contrast region and a region other than the contrast region (see the right diagram in FIG. 7).
  • optimal region division can be performed by minimizing the total cost of all t-links and n-links to be cut. That is, when the cost function of t-link is fv (Xv) and the cost function of n-link is fuv (Xu, Yv), the region is divided so that the total cost E (x) in the following equation is minimized.
  • the t-link that connects the node N ijk and the node T representing the region other than the myocardial region is, for example, a CT value around the myocardial region in which the pixel value is statistically acquired in advance. It can be calculated based on the determination result of whether or not it is within the range of values.
  • the t-link connecting the node N ijk and the node S representing the myocardial region is based on, for example, a determination result of whether or not the pixel value is statistically within the CT value range of the myocardial region acquired in advance. Can be calculated.
  • one of the two t-links connecting the node representing each pixel and the node S representing the myocardial region or the node T representing other than the myocardial region is cut, and nodes of adjacent pixels having different labels are connected to each other.
  • the n-link to be connected it is divided into a myocardial region and a region other than the myocardial region. Also at this time, it is possible to perform appropriate region division by minimizing the total cost of all t-links and n-links to be cut, that is, by obtaining an optimal solution.
  • the region extraction processing unit 12 extracts the contrast region and the myocardial region constituting the left ventricular region by the graph cut method, thereby obtaining the final left ventricular region LV that combines these regions. To do.
  • the region extraction processing unit 12 extracts the right ventricular region of the heart based on the extraction result of the left ventricular region as described above (S14).
  • the region extraction processing unit 12 first sets an initial region of the right ventricular region using atlas image data including a heart image, as in the case of extracting the left ventricular region.
  • the region extraction processing unit 12 uses the pixels in the initial region of the right ventricular region extracted as described above and the pixels around the initial region as extraction processing target pixels to generate a graphical model in the graph cut method. Generate. Then, the region extraction processing unit 12 extracts the final right ventricular region RV by performing region extraction processing using the graph cut method, and at this time, performs region extraction processing using the extraction result of the left ventricular region. .
  • the region extraction processing unit 12 sets the above-described cost function fv (Xv) (corresponding to the evaluation function) so as to assign a relatively large cost to the pixels extracted as the left ventricular region LV. This prevents the pixels extracted as the left ventricular region from being extracted as pixels of the right ventricular region.
  • the contents of the region extraction process by the graph cut method other than the setting of the cost function fv (Xv) described above are the same as the region extraction process of the left ventricular region described above.
  • the cost function for the pixels extracted as the left ventricular region is controlled as a method using the extraction result of the left ventricular region.
  • the present invention is not limited to this.
  • the above-described graphical model is created.
  • the graphical model may be created by excluding the pixels extracted as the left ventricular region.
  • the region extraction processing unit 12 extracts the left atrial region of the heart based on the extraction result of the left ventricular region and the right ventricular region as described above (S16).
  • the region extraction processing unit 12 sets the initial region of the left atrial region using the atlas image data in the same manner as when the right ventricular region is extracted.
  • the region extraction processing unit 12 sets a graphical model in the graph cut method using the pixels in the initial region of the left atrial region extracted as described above and the pixels around the initial region as extraction target pixels. Generate. Then, the region extraction processing unit 12 extracts a final left atrial region by performing region extraction processing using a graph cut method. At this time, region extraction is performed using the extraction results of the left ventricular region and the right ventricular region. Process.
  • the region extraction processing unit 12 uses the above-described cost function fv (Xv) so as to assign a relatively large cost to the pixels extracted as the left ventricular region and the pixels extracted as the right ventricular region.
  • the pixel extracted as the left ventricular region and the pixel extracted as the right ventricular region are not extracted as pixels of the left atrial region.
  • the contents of the region extraction process by the graph cut method other than the setting of the cost function fv (Xv) described above are the same as the region extraction process described above.
  • the graphical model when extracting the left atrial region, when creating the graphical model, may be created by excluding the pixels extracted as the left ventricular region and the pixels extracted as the right ventricular region. Good.
  • the region extraction processing unit 12 extracts the right atrial region of the heart based on the extraction results of the left ventricular region, the right ventricular region, and the left atrial region as described above (S18).
  • the region extraction processing unit 12 sets the initial region of the right atrial region using the atlas image data in the same manner as when the right ventricular region is extracted.
  • the region extraction processing unit 12 sets a graphical model in the graph cut method using the pixels in the initial region of the right atrial region extracted as described above and the pixels around the initial region as extraction target pixels. Generate. Then, the region extraction processing unit 12 extracts a final right atrial region by performing region extraction processing using a graph cut method. At this time, the extraction results of the left ventricular region, the right ventricular region, and the left atrial region are extracted. Use this to perform region extraction processing.
  • the region extraction processing unit 12 assigns a relatively large cost to the pixels extracted as the left ventricular region, the pixels extracted as the right ventricular region, and the pixels extracted as the left atrial region.
  • the cost function fv (Xv) described above the pixel extracted as the left ventricular region, the pixel extracted as the right ventricular region, and the pixel extracted as the left atrial region are not extracted as pixels of the right atrial region.
  • the contents of the region extraction process by the graph cut method other than the setting of the cost function fv (Xv) described above are the same as the region extraction process described above.
  • the region extraction processing unit 12 extracts the left ventricular region, the right ventricular region, the left atrial region, and the right atrial region in this order. Then, the region extraction processing unit 12 outputs the extraction result to the display control unit 13.
  • the display control unit 13 Based on the input extraction result, the display control unit 13 generates images color-coded in the left ventricle region LV, the right ventricular region RV, the left atrial region LA, and the right atrial region RA, as shown in FIG. Then, an image obtained by superimposing the generated color-coded image on the medical image is displayed on the display 20.
  • medical image data including a heart image is acquired, the left ventricle region of the heart in the acquired medical image data is independently extracted, and the extraction result of the left ventricular region is extracted. Since at least one of the right ventricular region, left atrial region, and right atrial region of the heart is extracted based on the above, each region of the heart can be extracted with high accuracy.
  • the shape of the left ventricle region of the heart is relatively simple, it can be extracted with high accuracy by extracting it alone. Then, by extracting the other region using the extraction result of the left ventricle region as reliable information, the extraction accuracy of the other region can be improved, and the extraction error can be reduced.
  • the right ventricle region is then extracted.
  • the right ventricular region is a region adjacent to the left ventricular region in the largest range. Therefore, by using the extraction result of the left ventricle region, it is possible to effectively prevent over-extraction to the left ventricle region side or insufficient extraction of the right ventricle region.
  • the left atrial region is extracted first among the left atrial region and the right atrial region.
  • the right atrial region may be extracted first. Good.
  • the extraction result of the left ventricular region and the right ventricular region is used, and when extracting the left atrial region, the left ventricular region, the right ventricular region, and the right atrial region are extracted.
  • the extraction result may be used.
  • the region extraction processing by the graph cut method is performed.
  • the region extraction processing is limited to this. It is not a thing.
  • the final right ventricular region may be extracted by performing region extraction processing using the atlas image data used when extracting the initial region in the above description.
  • non-rigid registration is performed between the atlas image data and the medical image data to be extracted to obtain the correspondence between the pixels of the two image data as a matching function. . Then, the pixels extracted as the left ventricle region at this time are excluded from the target pixels when the matching function is obtained.
  • region extraction processing using atlas image data may be performed in the same manner as described above.
  • the pixel extracted as the left ventricular region and the pixel extracted as the right ventricular region are excluded from the target pixels when obtaining the matching function
  • the right atrial region When extracting a pixel, a pixel extracted as the left ventricular region, a pixel extracted as the right ventricular region, and a pixel extracted as the left atrial region are excluded from the target pixels when obtaining the matching function. That's fine.
  • the region extraction processing using the dynamic contour model is performed. Also good.
  • an evaluation function is defined, and an optimal solution is obtained by calculating the evaluation function while changing the position of the contour point of the extraction target region.
  • Region extraction processing using an active contour model is a well-known method. For example, “Eito:“ Overview of Active Contour Model Snakes ”, Medical Imaging Technology, vol. 12, no.1, pp.9-15. , 1994 "and” M. Kass, A. Wilin, D.
  • Eint is a term indicating internal energy, and is a term for controlling the smoothness of the contour.
  • Eimage is a term indicating image energy, and is a term for controlling the position of the contour to a position where the luminance change is large.
  • Econ is a term indicating external energy. For example, when a part of the contour is to be fixed by external designation, the difference term between the point v (s) on the contour and the designated point p (v (s)- p) 2 should be added.
  • the evaluation value deteriorates when a contour point enters the left ventricular region as Econ, which is a term indicating external energy.
  • a penalty term may be provided. Thereby, extraction mistakes can be suppressed.
  • region extraction processing using an active contour model may be performed in the same manner as described above.
  • the evaluation value is deteriorated when a contour point enters the left ventricular region and the right ventricular region as Econ which is a term indicating external energy.
  • Econ which is a term indicating external energy
  • Econ which is a term indicating external energy
  • a penalty term that deteriorates the evaluation value may be provided.
  • the medical image data acquisition unit 11 acquires a plurality of medical image data having different phases in one cardiac cycle of the beating heart, and the left and right ventricular regions and the left and right atrial regions are obtained from the plurality of medical image data. May be extracted.
  • one piece of medical image data in the expansion period is selected from among a plurality of pieces of medical image data having different phases.
  • the medical image data in the end diastole phase can be said to be medical image data more suitable for region extraction in that the movement is small and the shape is stable.
  • medical image data in other phases in the diastole may be used.
  • the left ventricle region LV is first extracted by the same method as described above.
  • the right ventricular region RV, the left atrial region LA, and the right atrial region RA in the medical image data in the end diastole phase are extracted by the same method as described above.
  • the left and right ventricular regions LV and RV and the left and right atrial regions LA and RA in the phase of the medical image data other than the single medical image data are extracted.
  • the correspondence between the pixels of the two image data is obtained as a matching function.
  • the left and right ventricular regions LV and RV and the left and right atrial regions LA and RA on the medical image data in the end diastole phase other than the one medical image data are obtained, and these regions are extracted as left and right ventricular regions LV and RV and left and right atrial regions LA and RA.
  • the method for extracting the left and right ventricular regions LV and RV and the left and right atrial regions LA and RA of the medical image data in a phase other than the medical image data in the end diastole phase is not limited to the above method.
  • the extraction results of the left and right ventricular regions LV and RV and the left and right atrial regions LA and RA of the medical image data of the phase are used as initial regions of the left and right ventricular regions LV and RV and the left and right atrial regions LA and RA of the medical image data of the other phases.
  • the final left and right ventricular regions LV and RV and left and right atrial regions LA and RA may be extracted using the graph cut method described above.
  • the region extraction processing unit 12 accepts manual correction of the landmark position set when the above-described initial region is obtained, and the corrected landmark The initial region may be extracted again based on the position of the region, and the region extraction process described above may be performed again.
  • a correction acceptance screen on which medical image data before region extraction processing is displayed again is displayed on the display 20, and the user inputs settings using the input device 30. You can do it.
  • the region extraction processing unit 12 accepts manual correction of the contour point of the left ventricular region, and uses the left ventricular region after the correction to use the right ventricular region.
  • the left atrial region and the right atrial region may be extracted.
  • the manual correction of the contour point of the left ventricular region may be performed by the user using the input device 30 for setting.
  • the region extraction processing unit 12 manually performs the contour point of each region on the medical image data in the end diastole phase.
  • the correction may be received, and the right ventricular region, the left atrial region, and the right atrial region on the medical image data of a phase other than the single medical image data may be extracted using each region after the correction.
  • the user may set and input using the input device 30.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Radiology & Medical Imaging (AREA)
  • Molecular Biology (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Biomedical Technology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Veterinary Medicine (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Software Systems (AREA)
  • Cardiology (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)
PCT/JP2014/000076 2013-01-16 2014-01-10 医用画像処理装置および方法並びにプログラム Ceased WO2014112338A1 (ja)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US14/800,704 US10019804B2 (en) 2013-01-16 2015-07-16 Medical image processing apparatus, method, and program

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2013005050A JP6195714B2 (ja) 2013-01-16 2013-01-16 医用画像処理装置および方法並びにプログラム
JP2013-005050 2013-01-16

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US14/800,704 Continuation US10019804B2 (en) 2013-01-16 2015-07-16 Medical image processing apparatus, method, and program

Publications (1)

Publication Number Publication Date
WO2014112338A1 true WO2014112338A1 (ja) 2014-07-24

Family

ID=51209436

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2014/000076 Ceased WO2014112338A1 (ja) 2013-01-16 2014-01-10 医用画像処理装置および方法並びにプログラム

Country Status (3)

Country Link
US (1) US10019804B2 (enExample)
JP (1) JP6195714B2 (enExample)
WO (1) WO2014112338A1 (enExample)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5946221B2 (ja) * 2013-06-11 2016-07-05 富士フイルム株式会社 輪郭修正装置、方法およびプログラム
JP6411183B2 (ja) * 2014-11-13 2018-10-24 キヤノンメディカルシステムズ株式会社 医用画像診断装置、画像処理装置及び画像処理プログラム
JP6643827B2 (ja) 2015-07-31 2020-02-12 キヤノン株式会社 画像処理装置、画像処理方法、及びプログラム
JP6611660B2 (ja) 2016-04-13 2019-11-27 富士フイルム株式会社 画像位置合わせ装置および方法並びにプログラム
KR101831340B1 (ko) 2016-09-07 2018-02-23 서울여자대학교 산학협력단 복부 컴퓨터 단층촬영영상에서 다중 확률 아틀라스 기반 형상제한 그래프-컷을 사용한 신실질 자동 분할 방법 및 장치
JP6740910B2 (ja) 2017-01-13 2020-08-19 コニカミノルタ株式会社 動態画像処理システム
US11562487B2 (en) * 2017-10-18 2023-01-24 Koninklijke Philips N.V. Landmark visualization for medical image segmentation
JP6807981B2 (ja) * 2019-05-22 2021-01-06 富士フイルム株式会社 画像位置合わせ装置および方法並びにプログラム
CN110543912B (zh) * 2019-09-02 2021-10-01 李肯立 自动获取胎儿关键切面超声视频中心动周期视频的方法
WO2023193290A1 (zh) * 2022-04-08 2023-10-12 胡冠彤 面向体外心脏模拟器的医学成像系统和方法

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007044346A (ja) * 2005-08-11 2007-02-22 Toshiba Corp 医用画像処理装置における関心領域の経時的特定方法及び医用画像処理装置
WO2012153539A1 (ja) * 2011-05-11 2012-11-15 株式会社 東芝 医用画像処理装置とその方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6961454B2 (en) * 2001-10-04 2005-11-01 Siemens Corporation Research, Inc. System and method for segmenting the left ventricle in a cardiac MR image
US20090161926A1 (en) * 2007-02-13 2009-06-25 Siemens Corporate Research, Inc. Semi-automatic Segmentation of Cardiac Ultrasound Images using a Dynamic Model of the Left Ventricle
US8139838B2 (en) * 2008-05-22 2012-03-20 Siemens Aktiengesellschaft System and method for generating MR myocardial perfusion maps without user interaction
US8218839B2 (en) * 2008-05-23 2012-07-10 Siemens Aktiengesellschaft Automatic localization of the left ventricle in cardiac cine magnetic resonance imaging
WO2011083789A1 (ja) * 2010-01-07 2011-07-14 株式会社 日立メディコ 医用画像診断装置と医用画像の輪郭抽出処理方法
US8781552B2 (en) * 2011-10-12 2014-07-15 Siemens Aktiengesellschaft Localization of aorta and left atrium from magnetic resonance imaging

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007044346A (ja) * 2005-08-11 2007-02-22 Toshiba Corp 医用画像処理装置における関心領域の経時的特定方法及び医用画像処理装置
WO2012153539A1 (ja) * 2011-05-11 2012-11-15 株式会社 東芝 医用画像処理装置とその方法

Also Published As

Publication number Publication date
US20150317799A1 (en) 2015-11-05
JP6195714B2 (ja) 2017-09-13
US10019804B2 (en) 2018-07-10
JP2014135990A (ja) 2014-07-28

Similar Documents

Publication Publication Date Title
JP6195714B2 (ja) 医用画像処理装置および方法並びにプログラム
Noothout et al. Deep learning-based regression and classification for automatic landmark localization in medical images
Kirisli et al. Fully automatic cardiac segmentation from 3D CTA data: a multi-atlas based approach
JP6058093B2 (ja) 医療用画像のコンピュータ支援による解析装置、および、医療用画像解析のためのコンピュータプログラム
JP4918048B2 (ja) 画像処理装置及び方法
JP5347003B2 (ja) 医用画像処理装置および方法ならびにプログラム
JP5438029B2 (ja) 短軸遅延強調心臓ｍｒｉの自動三次元セグメント化
US8867822B2 (en) Model-based coronary artery calcium scoring
US9931790B2 (en) Method and system for advanced transcatheter aortic valve implantation planning
US8948484B2 (en) Method and system for automatic view planning for cardiac magnetic resonance imaging acquisition
Elattar et al. Automatic segmentation of the aortic root in CT angiography of candidate patients for transcatheter aortic valve implantation
JP5777070B2 (ja) 領域抽出装置、領域抽出方法および領域抽出プログラム
CN105793893B (zh) 解剖结构的基于模型的分割
CN102715906A (zh) 从c臂血管造影术的单扫描估算3d心脏运动的方法和系统
Fritz et al. Automatic segmentation of the left ventricle and computation of diagnostic parameters using regiongrowing and a statistical model
O’Donnell et al. Comprehensive cardiovascular image analysis using MR and CT at Siemens Corporate Research
Müller et al. Interventional heart wall motion analysis with cardiac C-arm CT systems
Zhao et al. Robust shape-constrained active contour for whole heart segmentation in 3-D CT images for radiotherapy planning
Chen et al. Automatic extraction of 3D dynamic left ventricle model from 2D rotational angiocardiogram
Medina et al. Left ventricle myocardium segmentation in multi-slice computerized tomography
Vera et al. Similarity enhancement for automatic segmentation of cardiac structures in computed tomography volumes
EP4607530A1 (en) Method for generating a personalized 4d cardiac model
Alvén et al. Überatlas: Robust speed-up of feature-based registration and multi-atlas segmentation
Meyer et al. Fully automatic segmentation of complex organ systems: example of trachea, esophagus and heart segmentation in CT images
Meyer et al. A multi-modality segmentation framework: application to fully automatic heart segmentation

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: 14740742

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14740742

Country of ref document: EP

Kind code of ref document: A1