Information processing method, program and information processor

Abstract

To provide an information processing method, etc. capable of presenting a detection processing range when detecting a detection target area from a blood vessel tomographic image.SOLUTION: An information processing method includes the following steps of: acquiring a blood vessel tomographic image; inputting the acquired blood vessel tomographic image to a model learned so as to detect a detection target area when inputting the blood vessel tomographic image, and thereby detecting the detection target area; and allowing the computer to perform processing for superimposing and displaying, on the blood vessel tomographic image, a detection result of the detection target area and a detection processing range by the model.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present invention relates to an information processing method, a program, and an information processing device. [Background technology]

[0002] There is a technology that uses machine learning to support image diagnosis related to vascular treatment such as PCI (Percutaneous Coronary Intervention). For example, Patent Literature 1 discloses a system that uses a neural network to perform segmentation of a region of interest from a polar image of a coronary artery. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Special Publication No. 2022-522960 Summary of the Invention [Problem to be solved by the invention]

[0004] In one aspect, an object of the present invention is to provide an information processing method and the like that can present a detection processing range when detecting a detection target region from a vascular tomographic image. [Means for solving the problem]

[0005] In one aspect, (1) an information processing method includes a computer executing a process to acquire a vascular tomographic image, input the acquired vascular tomographic image into a model that has been trained to detect a detection target area when a vascular tomographic image is input, thereby detecting the detection target area, and displaying the detection result of the detection target area and the detection processing range by the model superimposed on the vascular tomographic image.

[0006] (2) The information processing method of (1) above includes selecting a model that detects the detection target area from a plurality of models, and acquiring the detection processing range according to the selected model.

[0007] (3) In the information processing method of (2) above, the multiple models include at least two of a model trained to detect a detection target area of ​​the aorta, a model trained to detect a detection target area of ​​a coronary artery, a model trained to detect a detection target area of ​​a lower limb vein, and a model trained to detect a detection target area of ​​a lower limb artery.

[0008] (4) In the information processing method according to any one of (1) to (3) above, the detection processing range is determined according to a maximum depth of a vascular tomographic image used for learning the model.

[0009] (5) The information processing method described in any one of (1) to (4) above complements the boundary portion of the detection target area located outside the detection processing range based on the boundary portion of the detection target area detected within the detection processing range, and when the boundary portion of the detection target area located outside the detection processing range is complemented, the detection target area within the detection processing range and the detection target area outside the detection processing range are displayed in different display modes.

[0010] (6) In the information processing method described in any one of (1) to (5) above, the vascular tomographic image is a transverse cross-sectional image obtained by capturing a cross section intersecting the axial direction of the blood vessel, and a plurality of transverse cross-sectional images captured continuously along the axial direction are acquired, and each of the transverse cross-sectional images is input to the model to detect the detection target area from each of the transverse cross-sectional images. A longitudinal cross-sectional image of the blood vessel along the axial direction is generated based on the plurality of transverse cross-sectional images, and the detection result of the detection target area and the detection processing range are superimposed and displayed on the longitudinal cross-sectional image.

[0011] (7) In the information processing method of (6) above, when the detection target area detected by the model when the transverse section image is input includes an area that does not overlap with the detection processing range, the longitudinal section image is displayed in a different manner compared to when the detection target area falls within the detection processing range.

[0012] In one aspect, (8) the program causes a computer to execute a process of acquiring a vascular tomographic image, detecting the detection target area by inputting the acquired vascular tomographic image into a model that has been trained to detect a detection target area when a vascular tomographic image is input, and displaying the detection result of the detection target area and the detection processing range by the model superimposed on the vascular tomographic image.

[0013] In one aspect, (9) an information processing device is an information processing device that includes a control unit, wherein the control unit acquires a vascular tomographic image, detects the detection target area by inputting the acquired vascular tomographic image into a model that has been trained to detect a detection target area when a vascular tomographic image is input, and displays the detection result of the detection target area and the detection processing range by the model superimposed on the vascular tomographic image. Effect of the Invention

[0014] In one aspect, when a detection target region is detected from a vascular tomographic image, a detection processing range can be presented. [Brief description of the drawings]

[0015] [Figure 1] FIG. 1 is an explanatory diagram illustrating an example of the configuration of an image diagnostic system. [Diagram 2] FIG. 2 is a block diagram showing an example of the configuration of a server. [Diagram 3] FIG. 1 is a block diagram showing an example of the configuration of an image diagnostic apparatus. [Figure 4] FIG. 1 is an explanatory diagram of a detection model. [Diagram 5] FIG. 4 is an explanatory diagram relating to a detection processing range of a detection target region. [Figure 6]FIG. 13 is an explanatory diagram showing a display example of a transverse layer image. [Figure 7] FIG. 13 is an explanatory diagram showing a display example of a longitudinal tomographic image. [Figure 8] 13 is a flowchart showing a procedure for generating a detection model. [Figure 9] 10 is a flowchart showing a procedure for a detection process of a detection target area. [Figure 10] 13 is an explanatory diagram showing a display example of a longitudinal tomographic image according to a modified example. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] The present invention will now be described in detail with reference to the drawings showing embodiments thereof. (Embodiment) Fig. 1 is an explanatory diagram showing a configuration example of an image diagnostic system. In this embodiment, an image diagnostic system that detects a detection target area indicating a predetermined detection target (for example, a blood vessel lumen, a stent, an external elastic plate, etc.) from a vascular tomographic image obtained by capturing a cross section of a patient's blood vessel will be described. The image diagnostic system includes a server 1 and an image diagnostic device 2. Each device is communicatively connected via a network N.

[0017] The server 1 is a server computer capable of various information processing and transmitting and receiving information. The device equivalent to the server 1 may be a personal computer or other device. The server 1 performs machine learning to learn from predetermined training data, and generates a detection model 50 (see FIG. 4) that detects detection target regions such as the lumen, stent, and external elastic lamina of the blood vessel using a vascular tomographic image as input.

[0018] The image diagnostic device 2 is an imaging device that captures medical images of a patient's blood vessels, and is, for example, an IVUS (Intravascular Ultrasound) device that performs ultrasound examinations using a catheter 201. The catheter 201 is a medical instrument that is inserted into a patient's blood vessels, and transmits ultrasound signals from an ultrasound probe attached to the tip of the catheter 201 and receives reflected waves. The image diagnostic device 2 generates and displays an ultrasound tomographic image based on the reflected waves received by the catheter 201.

[0019] In this embodiment, the image diagnostic device 2 is described as an IVUS device, but it may be an optical imaging device using OCT (Optical Coherence Tomography). The image diagnostic device 2 may be a dual sensor system having both IVUS and OCT functions. The vascular tomographic image may be captured by a method other than IVUS or OCT.

[0020] Data of the detection model 50 generated by the server 1 is installed in the image diagnostic device 2, and the image diagnostic device 2 detects a detection target region by inputting a vascular tomographic image captured using a catheter 201 to the detection model 50. As will be described later, when the image diagnostic device 2 displays the detection result of the detection target region superimposed on the vascular tomographic image, it displays the detection processing range by the detection model 50 superimposed.

[0021] In this embodiment, the image diagnostic device 2 detects the detection target region using the detection model 50, but the server 1 on the cloud may execute the processing using the detection model 50. Also, for example, a general-purpose computer connected to the image diagnostic device 2 may execute the processing. In this way, the processing entity that executes the series of processes is not particularly limited.

[0022] 2 is a block diagram showing an example of the configuration of the server 1. The server 1 includes a control unit 11, a main storage unit 12, a communication unit 13, and an auxiliary storage unit . The control unit 11 has one or more arithmetic processing devices such as a central processing unit (CPU), a micro-processing unit (MPU), a graphics processing unit (GPU), etc., and performs various information processing, control processing, etc. by reading and executing a program P1 stored in the auxiliary storage unit 14. The main storage unit 12 is a temporary storage area such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), and temporarily stores data required for the control unit 11 to execute arithmetic processing. The communication unit 13 is a communication module for performing processing related to communication, and transmits and receives information to and from the outside. The auxiliary storage unit 14 is a non-volatile storage area such as a large-capacity memory or a hard disk, and stores the program P1 (program product) and other data required for the control unit 11 to execute processing.

[0023] The auxiliary storage unit 14 may be an external storage device connected to the server 1. The server 1 may be a multi-computer consisting of a plurality of computers, or may be a virtual machine virtually constructed by software.

[0024] In the present embodiment, the server 1 is not limited to the above configuration, and may include, for example, an input unit for accepting operation input, a display unit for displaying images, etc. The server 1 may also include a reading unit for reading a portable storage medium 1a such as a CD (Compact Disk)-ROM or a DVD (Digital Versatile Disc)-ROM, and may read and execute the program P1 from the portable storage medium 1a.

[0025] 3 is a block diagram showing an example of the configuration of the image diagnostic apparatus 2. The image diagnostic apparatus 2 includes a control unit 21, a main storage unit 22, a communication unit 23, a display unit 24, an input unit 25, an image processing unit 26, and an auxiliary storage unit 27. The control unit 21 has one or more arithmetic processing devices such as CPUs, and performs various information processing, control processing, etc. by reading and executing a program P2 stored in the auxiliary storage unit 27. The main storage unit 22 is a temporary storage area such as RAM, and temporarily stores data necessary for the control unit 21 to execute arithmetic processing. The communication unit 23 is a communication module for performing processing related to communication, and transmits and receives information to and from the outside. The display unit 24 is a display screen such as a liquid crystal display, and displays images. The input unit 25 is an operation interface such as a keyboard or mouse, and accepts operation input from the user. The image processing unit 26 is an image processing module that processes signals transmitted and received via the catheter 201, and generates a vascular tomographic image.

[0026] The auxiliary storage unit 27 is a non-volatile storage area such as a hard disk or a large-capacity memory, and stores a program P2 (program product) and other data necessary for the control unit 21 to execute processing. The auxiliary storage unit 27 also stores a first detection model 50a, a second detection model 50b, a third detection model 50c, and a fourth detection model 50d. The first detection model 50a, the second detection model 50b, the third detection model 50c, and the fourth detection model 50d are each a machine learning model generated by learning predetermined training data, and are trained models that detect a detection target region using a vascular tomographic image as an input.

[0027] The image diagnostic apparatus 2 may include a reading unit for reading a portable storage medium 2a such as a CD-ROM, and may read and execute the program P2 from the portable storage medium 2a.

[0028] The first detection model 50a, the second detection model 50b, the third detection model 50c, and the fourth detection model 50d are different in the types of blood vessels related to the blood vessel tomographic images that are the learning target. Specifically, the first detection model 50a is a model that has learned a tomographic image of the aorta, the second detection model 50b is a model that has learned a tomographic image of the coronary artery, the third detection model 50c is a model that has learned a tomographic image of the lower limb vein, and the fourth detection model 50d is a model that has learned a tomographic image of the lower limb artery. In this way, the types of blood vessels related to the learned blood vessel tomographic images are different for each model, and the detection processing range is different depending on the model, as described later.

[0029] In the following description, these four models are collectively referred to as "detection models 50."

[0030] Fig. 4 is an explanatory diagram of the detection model 50. Fig. 4 illustrates how, when a vascular tomographic image (transverse image) is input to the detection model 50, a detection result is output in which detection target regions such as the vascular lumen, stent, and external elastic membrane (EEM) are detected.

[0031] In this embodiment, the detection model 50 is described as detecting a detection target area from a transverse slice image capturing a cross section intersecting (orthogonal to) the axial direction of the blood vessel, but the detection target area may also be detected from a longitudinal slice image showing a cross section along the axial direction of the blood vessel.

[0032] The detection model 50 is a machine learning model that has learned predetermined training data, and when a vascular tomographic image is input, the detection model 50 detects a detection target region indicating one or more detection targets. For example, the detection model 50 is a model related to semantic segmentation such as U-net (CNN: Convolutional Neural Network), and detects the detection target region in pixel units.

[0033] The detection model 50 may be a neural network other than the CNN. The detection model 50 may be a machine learning model other than the neural network, such as a decision tree or a Support Vector Machine (SVM).

[0034] The detection target regions include a region showing the lumen of a blood vessel (main trunk), a region showing the inside of a stent, a region showing the inside of the external elastic lamina of a blood vessel (main trunk), a region showing the inside of the adventitia of a blood vessel, a region showing a side branch of a blood vessel, a plaque region, a thrombus region, a hematoma region, a calcification region, a dissection region, a crack region, an extravascular tissue region, and a device region, etc. Note that "device" refers to a device such as a guidewire that is inserted into a patient's blood vessel.

[0035] The above is an example of the detection target area, and other detection targets may be detected. The detection model 50 according to the present embodiment detects the detection target area indicating each of these multiple detection targets. As an example, FIG. 4 illustrates a case where the lumen, stent, and external elastic lamina of a blood vessel are detected.

[0036] In addition, in FIG. 4, the detection target area is illustrated as the inner area of ​​the object (for example, the area inside the lumen of a blood vessel), but this embodiment is not limited to this, and the detection model 50 may detect only the contour (boundary) portion of the object as the detection target area.

[0037] The server 1 generates a detection model 50 using training data to which a label indicating the position of the correct detection target region is added to a training vascular tomographic image. The server 1 detects the detection target region by inputting the training vascular tomographic image to the detection model 50. The server 1 adjusts parameters such as weights between neurons so that the detected detection target region approximates the correct label. The server 1 sequentially supplies a group of training vascular tomographic images to the detection model 50 to perform learning, and finally generates a detection model 50 with optimized parameters.

[0038] In this embodiment, the server 1 generates a plurality of detection models 50 according to the type of blood vessels related to the vascular tomographic image to be processed. Specifically, as already mentioned, the server 1 generates a first detection model 50a for the aorta, a second detection model 50b for the coronary artery, a third detection model 50c for the lower limb vein, and a fourth detection model 50d for the lower limb artery.

[0039] The first detection model 50a is a model generated based on training data using a tomographic image of the aorta as a training image, and preferably detects a detection target area from the tomographic image of the aorta. The second detection model 50b is a model generated based on training data using a tomographic image of the coronary artery as a training image, and preferably detects a detection target area from the tomographic image of the coronary artery. The third detection model 50c is a model generated based on training data using a tomographic image of the lower limb vein as a training image, and preferably detects a detection target area from the tomographic image of the lower limb vein. The fourth detection model 50d is a model generated based on training data using a tomographic image of the lower limb artery as a training image, and preferably detects a detection target area from the tomographic image of the lower limb artery.

[0040] The server 1 generates each model using different training data according to each model. That is, the server 1 generates a first detection model 50a using training data in which a label indicating a detection target region is added to a tomographic image of an aorta. The server 1 generates a second detection model 50b using training data in which a label indicating a detection target region is added to a tomographic image of a coronary artery. The server 1 generates a third detection model 50c using training data in which a label indicating a detection target region is added to a tomographic image of a lower limb vein. The server 1 generates a fourth detection model 50d using training data in which a label indicating a detection target region is added to a tomographic image of a lower limb artery.

[0041] When actually detecting a detection target region from a vascular tomographic image of a patient, the image diagnostic device 2 selects one of the detection models 50 according to the type of blood vessel related to the vascular tomographic image captured by the catheter 201. Then, the image diagnostic device 2 detects the detection target region by inputting into the selected detection model 50. The detection model may be selected automatically by image processing, or may be selected manually by the user via an interface.

[0042] The image diagnostic device 2 displays a vascular tomographic image on which the detection result of the detection target region is superimposed. For example, the image diagnostic device 2 displays a vascular tomographic image on which the boundary line of the detection target region is superimposed (see FIG. 6).

[0043] Here, in this embodiment, the image diagnostic device 2 displays not only the detection result of the detection target region, but also a vascular tomographic image on which the detection processing range by the detection model 50 is superimposed.

[0044] Fig. 5 is an explanatory diagram of the detection processing range of the detection target region. In Fig. 5, the detection processing range in which the detection result is highly reliable is illustrated by a thick dotted line in a blood vessel tomographic image.

[0045] There are two types of depth (distance from the center of the image) in vascular tomographic images: acquisition depth and display depth. Display depth refers to the maximum distance from the center of the image displayed on the console screen. Acquisition depth refers to the maximum depth acquired by ultrasound (or light in the case of OCT).

[0046] During endovascular treatment, a display depth that is shallower than the acquisition depth is often used. For example, for the coronary arteries, a depth of about 4 mm is sufficient. On the other hand, in the lower extremities, the blood vessels are thicker, so a larger range must be visualized for confirmation.

[0047] Although a method of training the detection model 50 on images of various depths (sizes) as data augmentation is possible, it is desirable to train the detection model 50 on a standardized image size because normalizing image conditions generally results in higher detection accuracy. For example, when training the detection model 50 for the purpose of PCI, it is desirable to train the model on a standardized image size of 4 mm.

[0048] In this case, when the user (medical worker) changes the display depth, an area beyond the image size learned by the detection model 50 will be displayed. For example, in the case of a large blood vessel, the blood vessel may be located deeper than 4 mm, but since this range is not within the original detection range, it will not be processed.

[0049] Therefore, in this embodiment, the detection processing range by the detection model 50 is clearly indicated as shown in Fig. 5. Specifically, the image diagnostic device 2 displays the boundary line of the detection processing range superimposed on the vascular tomographic image.

[0050] The detection processing range is determined according to the maximum depth (maximum display depth) of the training vascular tomographic image used for learning the detection model 50. That is, the image diagnostic device 2 sets the maximum depth of the training vascular tomographic image as the detection processing range. The image diagnostic device 2 stores the detection processing range according to the size of the image used during learning for each of the first detection model 50a for the aorta, the second detection model 50b for the coronary artery, the third detection model 50c for the lower limb vein, and the fourth detection model 50d for the lower limb artery. When the image diagnostic device 2 selects any one of the detection models 50 to detect the detection target region, it obtains the detection processing range according to the selected detection model 50 and displays the detection processing range superimposed on the vascular tomographic image together with the detection result of the detection target region.

[0051] Fig. 6 is an explanatory diagram showing a display example of a transverse section image. Figs. 6A and 6B show a display example of a transverse section image (vascular tomographic image) in which the detection result of the detection target region and the detection processing range of the detection result are superimposed. As shown in Fig. 6A, when the detection target region falls within the detection processing range, the image diagnostic device 2 displays the boundaries of each detection target region (lumen, stent, external elastic plate, etc.) with solid lines and displays the boundaries of the detection processing range with dotted lines. The detection processing range shown by the dotted lines is automatically changed (enlarged or reduced) according to the change in the display depth by the user.

[0052] When the detection target area does not fall within the detection processing range and extends outside the detection processing range, the image diagnostic device 2 complements the boundary portion of the detection target area located outside the detection processing range by spline complementation or the like based on the boundary portion of the detection target area within the detection processing range. Then, the image diagnostic device 2 displays the detection target area within the detection processing range and the detection target area outside the detection processing range in different display modes. Specifically, as shown in FIG. 6B, the image diagnostic device 2 displays the boundary portion of the detection target area within the detection processing range with a solid line and the boundary portion of the detection target area outside the detection processing range with a dotted line. In the example of FIG. 6B, since the boundary of the external elastic lamina of the blood vessel is partially located outside the detection processing range, the boundary portion is displayed with a dotted line. In this way, by making the display mode different between the detection processing range and the outside of the detection processing range, the part detected by the detection model 50 can be presented to the user.

[0053] Fig. 7 is an explanatory diagram showing a display example of a longitudinal tomographic image. Although the case of displaying a transverse tomographic image has been described in Fig. 6, as shown in Fig. 7, the image diagnostic device 2 may generate a longitudinal tomographic image from a plurality of transverse tomographic images along the axial direction of the blood vessel, and display the detection result of the detection target region and the detection processing range in a superimposed manner.

[0054] In the longitudinal tomographic image illustrated in FIG. 7, the different display modes (e.g., display colors) of parts of the vascular wall are illustrated by hatching. For example, the image diagnostic device 2 detects a plaque region as a detection target region, and changes the display color of the vascular wall according to the ratio of the detected plaque to the vascular cross section. The ratio is calculated, for example, by (area of ​​the plaque region) / (area of ​​the plaque region+area of ​​the region showing the lumen of the blood vessel). The image diagnostic device 2 displays the vascular wall at each position along the axial direction on the longitudinal tomographic image in a different display color according to the calculated ratio. For example, the image diagnostic device 2 displays the vascular wall portion where the calculated ratio is equal to or greater than a threshold in a display color different from the other portions.

[0055] Furthermore, the image diagnostic device 2 displays a vertical bar marker 71 at each position of the blood vessel along the axial direction. The marker 71 indicates the presence or absence of a side branch. When a side branch is detected as a detection target region from the transverse slice image, the image diagnostic device 2 displays the marker 71 at a position on the longitudinal slice corresponding to the transverse slice image.

[0056] Furthermore, the image diagnostic device 2 superimposes the detection processing range on the longitudinal tomographic image. For example, the image diagnostic device 2 superimposes the boundary line (dotted line) of the detection processing range as shown in Fig. 7. This makes it possible to present to the user the range where the detection result is highly reliable, similar to the transverse tomographic image.

[0057] Fig. 8 is a flowchart showing the procedure of the generation process of the detection model 50. The process content when generating the detection model 50 by machine learning will be described with reference to Fig. 8. The control unit 11 of the server 1 acquires training data for generating the detection model 50 (step S11). The training data is data in which a label indicating the position of the correct detection target region is added to a training vascular tomographic image.

[0058] In this embodiment, the control unit 11 acquires training data for generating a first detection model 50a for detecting a detection target region of the aorta, training data for generating a second detection model 50b for detecting a detection target region of the coronary artery, training data for generating a third detection model 50c for detecting a detection target region of the lower limb vein, and training data for generating a fourth detection model 50d for detecting a detection target region of the lower limb artery. The training data for generating the first detection model 50a is data in which a label indicating the position of the correct detection target region is added to a cross-sectional image of the aorta. The training data for generating the second detection model 50b is data in which a label indicating the position of the correct detection target region is added to a cross-sectional image of the coronary artery. The training data for generating the third detection model 50c is data in which a label indicating the position of the correct detection target region is added to a cross-sectional image of the lower limb vein. The training data for generating the fourth detection model 50d is data in which a label indicating the position of the correct detection target region is added to a cross-sectional image of the lower limb artery.

[0059] The control unit 11 generates a detection model 50 that has been trained to detect a detection target region when a vascular tomographic image is input based on the training data (step S12). For example, the control unit 11 generates a CNN related to semantic segmentation as the detection model 50. The control unit 11 generates the first detection model 50a, the second detection model 50b, the third detection model 50c, and the fourth detection model 50d using training data for generating the first detection model 50a, the training data for generating the second detection model 50b, the training data for generating the third detection model 50c, and the training data for generating the fourth detection model 50d, respectively.

[0060] The control unit 11 determines the detection processing range of each model according to the maximum depth of the training vascular tomographic image used for learning each model (step S13). The control unit 11 stores (installs) the data of the first detection model 50a, the second detection model 50b, and the third detection model 50c in the image diagnostic device 2 in association with the detection processing range according to each model (step S14). The control unit 11 ends the series of processes.

[0061] 9 is a flowchart showing the procedure of the detection process of the detection target region. The process of detecting the detection target region from the vascular tomographic image using the detection model 50 and displaying the detection result will be described with reference to FIG. The control unit 21 of the image diagnostic device 2 acquires a vascular tomographic image obtained by capturing a cross section of a patient's blood vessel (step S31). Specifically, the control unit 21 acquires a plurality of transverse slice images captured continuously along the axial direction of the blood vessel.

[0062] The control unit 21 selects one of the detection models 50 (the first detection model 50a, the second detection model 50b, the third detection model 50c, and the fourth detection model 50d) according to the type of blood vessel related to the acquired vascular tomographic image (step S32). The control unit 21 acquires a detection processing range according to the selected detection model 50 (step S33).

[0063] The control unit 21 detects the detection target region by inputting the vascular tomographic image to the detection model 50 selected in step S32 (step S34). Specifically, the control unit 21 detects the detection target region from each transverse cross-sectional image by inputting each transverse cross-sectional image acquired in step S31 to the detection model 50.

[0064] The control unit 21 generates a longitudinal image of the blood vessel along the axial direction based on the multiple transverse images acquired in step S31 (step S35).

[0065] The control unit 21 displays a vascular tomographic image in which the detection result of the detection target region detected in step S34 and the detection processing range by the detection model 50 are superimposed (step S36). Specifically, when displaying a transverse cross-sectional image, the control unit 21 superimposes the boundary line of the detection target region and displays the transverse cross-sectional image in which the boundary line of the detection processing range is superimposed. Here, when the detection target region extends outside the detection processing range, the control unit 21 complements the boundary line of the detection target region located outside the detection processing range based on the boundary line of the detection target region within the detection processing range, and displays the detection target region within the detection processing range and the detection target region outside the detection processing range in different display modes. For example, the control unit 21 displays the boundary line of the detection target region outside the detection processing range with a dotted line.

[0066] Furthermore, when displaying a longitudinal tomographic image, the control unit 21 superimposes the boundary line of the detection processing range on the longitudinal tomographic image in which the display color of the vascular wall is changed according to the amount of plaque, and the marker 71 is added according to the presence or absence of a side branch. The control unit 21 ends the series of processes.

[0067] As described above, according to this embodiment, when a detection target region is detected from a vascular tomographic image, a detection processing range can be presented.

[0068] (Modification) In this modified example, when a longitudinal tomographic image is displayed, if the detection target area includes an area that does not overlap with the detection processing range, the longitudinal tomographic image is displayed in a different manner compared to when the detection target area falls within the detection processing range.

[0069] Fig. 10 is an explanatory diagram showing a display example of a longitudinal tomographic image according to a modified example, in which a box-shaped dotted frame is superimposed on a portion where a detection target region is detected from outside the detection processing range.

[0070] As described above, the image diagnostic device 2 detects a detection target region from each transverse cross-sectional image by inputting each transverse cross-sectional image captured along the axial direction of the blood vessel to the detection model 50. In this modification, when the detection target region is located outside the detection processing range, the image diagnostic device 2 displays the detection target region at a position on the longitudinal tomographic image corresponding to the transverse cross-sectional image located outside the detection processing range, differently from other regions.

[0071] Specifically, as shown in Fig. 10, the image diagnostic device 2 superimposes a dotted frame indicating that the detection target region is located outside the detection processing range. In the region on the right side of the longitudinal tomographic image in Fig. 10, blood vessels are widely spread, and the lumen and external elastic lamina of the blood vessels are located outside the detection processing range, so the dotted frame is superimposed. This makes it possible to present the region outside the detection processing range to the user.

[0072] 10, the detection target region is indicated by a dotted frame to indicate that it is located outside the detection processing range, but this embodiment is not limited to this, and for example, the border of the blood vessel outside the detection processing range may be displayed by a dotted line, or the display color of the inside of the blood vessel wall may be changed. In this way, it is sufficient for the image diagnostic device 2 to clearly indicate the region outside the detection processing range.

[0073] As described above, in this modification, when the detection target region detected by the detection model 50 includes a region that does not overlap with the detection processing range, the image diagnostic device 2 displays the longitudinal tomographic image in a different manner compared to when the detection target region falls within the detection processing range. Since the other points are the same as those in the embodiment, detailed explanations of the flowchart and other details will be omitted in this modification.

[0074] The embodiments disclosed herein are illustrative in all respects and should not be considered as limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the scope and meaning equivalent to the claims.

[0075] The matters described in each embodiment can be combined with each other. In addition, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations regardless of the citation format. Furthermore, the claims use a format in which a claim cites two or more other claims (multi-claim format), but this is not limited to this. They may also be written using a format in which a multiple claim cites at least one other multiple claim (multi-multi claim). [Explanation of symbols]

[0076] 1 Server 11 Control section 12 Main memory 13. Communications Department 14 Auxiliary storage P1 Program 2. Diagnostic imaging equipment 21 Control section 22 Main memory 23 Communications Department 24 Display section 25 Input section 26 Image Processing Section 27 Auxiliary storage P2 Program 50 Detection Models

Claims

1. Acquire a vascular tomographic image, A model that has been trained to detect a detection target region when a vascular tomographic image is inputted is provided, and the detection target region is detected by inputting the acquired vascular tomographic image. The detection result of the detection target region and the detection processing range by the model are displayed superimposed on the blood vessel tomographic image. An information processing method in which processing is performed by a computer.

2. selecting a model that detects the detection target region from the plurality of models; Obtaining the detection processing range according to the selected model The information processing method according to claim 1 .

3. The plurality of models includes at least two of a model trained to detect a detection target region of an aorta, a model trained to detect a detection target region of a coronary artery, a model trained to detect a detection target region of a lower limb vein, and a model trained to detect a detection target region of a lower limb artery. The information processing method according to claim 2 .

4. The detection processing range is determined according to the maximum depth of the vascular tomographic image used for learning the model. The information processing method according to claim 1 .

5. complementing a boundary portion of the detection target area located outside the detection processing range based on a boundary portion of the detection target area detected within the detection processing range; When a boundary portion of the detection target area located outside the detection processing range is complemented, the detection target area within the detection processing range and the detection target area outside the detection processing range are displayed in different display modes. The information processing method according to claim 1 .

6. The vascular tomographic image is a transverse cross-sectional image obtained by capturing a cross section intersecting an axial direction of a blood vessel, Acquiring a plurality of transverse slice images successively taken along the axial direction; Detecting the detection target region from each of the transverse slice images by inputting each of the transverse slice images into the model; generating a longitudinal image of the blood vessel along the axial direction based on the plurality of transverse images; The detection result of the detection target area and the detection processing range are displayed in a superimposed manner on the longitudinal tomographic image. The information processing method according to claim 1 .

7. When the detection target area detected by the model when the transverse slice image is input includes an area that does not overlap with the detection processing range, the longitudinal slice image is displayed in a different manner compared to when the detection target area falls within the detection processing range. The information processing method according to claim 6.

8. Acquire a vascular tomographic image, A model that has been trained to detect a detection target region when a vascular tomographic image is inputted is provided, and the detection target region is detected by inputting the acquired vascular tomographic image. The detection result of the detection target region and the detection processing range by the model are displayed superimposed on the blood vessel tomographic image. A program that causes a computer to carry out processing.

9. An information processing device including a control unit, The control unit: Acquire a vascular tomographic image, A model that has been trained to detect a detection target region when a vascular tomographic image is inputted is provided, and the detection target region is detected by inputting the acquired vascular tomographic image. The detection result of the detection target region and the detection processing range by the model are displayed superimposed on the blood vessel tomographic image. Information processing device.

Images