Arrhythmia type estimation device, arrhythmia type estimation method, arrhythmia type estimation program

Abstract

This system accurately estimates the type of tachyarrhythmia from images of the heart. [Solution] The arrhythmia type estimation device (1) includes an acquisition unit (101) that acquires a target signal waveform showing the time-series change in the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing an image of the subject's heart, and an estimation unit (105) that estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as the normal range.

Description

[Technical Field]

[0001] The present invention relates to a technique for determining the type of arrhythmia. [Background Art]

[0002] Techniques for grasping the relationship between the left and right sides of the heart and the relationship between the atria and ventricles from ultrasonic diagnostic moving images of the heart are widely used. For example, Patent Document 1 discloses an ultrasonic diagnostic apparatus that identifies boundary position groups in a plurality of heart chambers in an ultrasonic moving image, and acquires boundary positions of the plurality of heart chambers over a section of at least one heartbeat based on a tracking result of the boundary position groups. [Prior Art Literature] [Patent Literature]

[0003] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2022-149097 [Summary of the Invention] [Problem to be Solved by the Invention]

[0004] Conventionally, electrocardiography is mainly used for arrhythmia diagnosis, but there are cases where it is difficult to determine the type of occurring tachyarrhythmia, and there has been room for improvement in terms of determination accuracy.

[0005] One aspect of the present invention aims to accurately estimate the type of tachyarrhythmia from images obtained by photographing the heart. [Means for Solving the Problem]

[0006] An arrhythmia type estimation device according to one aspect of the present invention comprises: an acquisition unit that acquires a target signal waveform showing the time-series change in the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing an image of the heart of a subject; and an estimation unit that estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as the normal range.

[0007] An arrhythmia type estimation method according to one aspect of the present invention is an arrhythmia type estimation method performed by one or more information processing devices, comprising: an acquisition step of acquiring a target signal waveform that shows the time-series change of the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing an image of the heart of a subject; and an estimation step of estimating the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as a normal range.

[0008] Each aspect of the present invention may be implemented by a computer. In this case, a control program for the arrhythmia type estimation device, which enables the computer to implement the arrhythmia type estimation device by operating the computer as each part (software element) of the arrhythmia type estimation device, and a computer-readable recording medium on which the program is recorded, also fall within the scope of the present invention. [Effects of the Invention]

[0009] According to one aspect of the present invention, the type of tachyarrhythmia can be accurately estimated from images of the heart. [Brief explanation of the drawing]

[0010] [Figure 1] This figure shows an example configuration of an arrhythmia type estimation system according to one embodiment of the present invention. [Figure 2] This figure shows an example of detecting the left atrium, left ventricle, right atrium, and right ventricle regions from a cardiac ultrasound image. [Figure 3] This is a block diagram showing an example of the main components of an arrhythmia type estimation device according to Embodiment 1 of the present invention. [Figure 4] This is a diagram illustrating an example of the processing performed by the estimation unit. [Figure 5] This flowchart shows an example of the processing flow performed by an arrhythmia type estimation device. [Figure 6] This is a block diagram showing an example of the main components of an arrhythmia type estimation device according to Embodiment 2 of the present invention. [Figure 7] This diagram illustrates an example of the processing performed by the single-beat waveform acquisition unit. [Figure 8] This diagram illustrates an example of the processing performed by the single-beat waveform acquisition unit. [Figure 9] This is a diagram illustrating an example of the first pretreatment process. [Figure 10] This is a diagram illustrating an example of the processing performed by the estimation unit. [Figure 11] This flowchart shows an example of the processing flow performed by an arrhythmia type estimation device. [Figure 12] This is a block diagram showing an example of the main components of an arrhythmia type estimation device according to Embodiment 3 of the present invention. [Figure 13] This diagram illustrates an example of the processing performed by the second preprocessing unit and the estimation unit. [Figure 14] This flowchart shows an example of the processing flow performed by an arrhythmia type estimation device. [Modes for carrying out the invention]

[0011] [Embodiment 1] Embodiments of the present invention will be described in detail below.

[0012] (Overview of arrhythmia type estimation device 1) An arrhythmia type estimation apparatus 1 according to an embodiment of the present invention estimates the type of tachyarrhythmia that a subject is suffering from, based on the shape of a target signal waveform generated by analyzing an image of the heart of the subject. Here, the target signal waveform is data indicating a time-series change in the area of at least any one of the left atrium, left ventricle, right atrium, and right ventricle in an image of the heart of the subject. Hereinafter, the area of each region corresponding to the left atrium, left ventricle, right atrium, and right ventricle is referred to as "the area of each region".

[0013] The inventors have found that the shape of a signal waveform indicating a time-series change in the area of at least any one of the left atrium, left ventricle, right atrium, and right ventricle included in an image captured of the heart has features that correspond to the type of tachyarrhythmia occurring in the heart. The arrhythmia type estimation apparatus 1 can accurately estimate the type of tachyarrhythmia that the subject is suffering from from an image captured of the heart of the subject.

[0014] Tachyarrhythmia is an arrhythmia having a frequency in a frequency band higher than a reference frequency band predetermined as a normal range (for example, 100 bpm or more and 170 bpm or less). Tachyarrhythmias include, for example, atrial flutter (hereinafter referred to as AFL) and supraventricular tachycardia (hereinafter referred to as SVT). AFL is an arrhythmia in which fine movements of about 200 to 350 bpm occur in the atrium, and SVT is an arrhythmia in which the movements of the atrium and ventricle become faster than the frequency in the reference frequency band.

[0015] The subject may be an animal having a heart or a fetus thereof. For example, the subject may be a human fetus. In this case, the area of each atrium and each ventricle may be calculated based on an ultrasound image including a plurality of frame images captured of the fetus. An ultrasound image captured of the fetus includes the image of the fetus's heart, and the area of each region of the fetus's heart can be calculated from the ultrasound image. The process of calculating the area of each region of the heart from the ultrasound image will be described later with a specific example.

[0016] (Configuration of Arrhythmia Type Estimation System 100) First, the configuration of an arrhythmia type estimation system 100, including an arrhythmia type estimation device 1 according to one embodiment of the present invention, will be described with reference to Figure 1. Figure 1 is a diagram showing an example configuration of the arrhythmia type estimation system 100. The arrhythmia type estimation system 100 may include an imaging device 2, an image analysis device 3, an arrhythmia type estimation device 1, and a display device 4.

[0017] [Imaging device 2] The imaging device 2 may be an ultrasound imaging device capable of non-invasively imaging the inside of the subject's body. That is, the input image captured by the imaging device 2 may be an ultrasound image such as an echo image for tomographic imaging. The imaging device 2 may be connected to the image analysis device 3 in a communicative manner, as shown in the figure, in which case the image analysis device 3 may directly acquire the input image from the imaging device 2. The input image may be stored in an image management device (not shown) in association with subject information for each subject (e.g., electronic medical record information), in which case the arrhythmia type estimation device 1 only needs to acquire the input image from the image management device.

[0018] [Image analysis device 3] Image analysis device 3 detects the left atrium, left ventricle, right atrium, and right ventricle regions of the subject's heart as captured in the input image. Image analysis device 3 also calculates the area of ​​each detected region and generates a target signal waveform that shows the time-series change in the area of ​​each region.

[0019] The image analysis device 3 may be installed in the facility where the imaging device 2 is installed (for example, a medical facility), or it may be installed in a remote location. If installed in a remote location, the image analysis device 3 can acquire input images via communication over a communication network such as the Internet.

[0020] Here, the process of detecting the regions corresponding to each atrium and ventricle of the heart from the input image will be explained based on Figure 2. Figure 2 shows an example of detecting the regions of the left atrium, left ventricle, right atrium, and right ventricle from an ultrasound image of the heart. Image A1 in Figure 2 is an ultrasound image of the heart, and image A2 shows the detection results for each region R1 to R4 superimposed on the ultrasound image.

[0021] In image A1, the heart is visible in the area slightly below the center, and its external shape and the division of its interior into multiple compartments can be seen. Thus, in ultrasound images, each of the two ventricles and each of the two atria of the heart can be seen as a closed region. Therefore, by analyzing ultrasound images of the heart, it is possible to detect the left atrium, left ventricle, right atrium, and the respective regions R1-R4 of the right ventricle.

[0022] For example, machine learning can be performed using training data in which regions corresponding to the left atrium, left ventricle, right atrium, and right ventricle of the heart, as seen in cardiac ultrasound images, are labeled as ground truth data. These labels can also be called annotations. Through such machine learning, it is possible to construct a learning model capable of detecting the regions R1-R4 of the left atrium, left ventricle, right atrium, and right ventricle from cardiac ultrasound images. For example, by constructing a learning model using a convolutional neural network and using that model, highly accurate region detection becomes possible.

[0023] Image A2 in Figure 2 shows the results of detection using such a learning model. The detected regions R1 to R4 are the right ventricle, left ventricle, left atrium, and right atrium, respectively. By detecting each region R1 to R4, it becomes possible to calculate the area of ​​each region R1 to R4. For example, the area of ​​each region R1 to R4 can also be represented by the number of pixels contained in each detected region R1 to R4. By calculating the area of ​​each region R1 to R4 of the left atrium, left ventricle, right atrium, and right ventricle of the subject's heart in each frame of the ultrasound image, it is possible to obtain a target signal waveform that shows the time-series change of the area of ​​each region R1 to R4.

[0024] [Arrhythmia Type Estimation Device 1] The arrhythmia type estimation device 1 uses the target signal waveform acquired from the image analysis device 3 to estimate the type of tachyarrhythmia the subject is suffering from. Compared to estimating the type of tachyarrhythmia using an electrocardiogram, the arrhythmia type estimation device 1 can output more accurate estimation results.

[0025] The arrhythmia type estimation device 1 may be connected to the image analysis device 3 in a communicative manner, as shown in the figure. In this case, the arrhythmia type estimation device 1 may directly acquire the target signal waveform from the image analysis device 3. The target signal waveform generated by the image analysis device 3 may be stored in a portable recording medium, for example. In this case, the arrhythmia type estimation device 1 can read the target signal waveform from the recording medium. Alternatively, the target signal waveform generated by the image analysis device 3 may be stored in an arbitrary storage device (not shown) in association with the target subject information for each target subject. In this case, the arrhythmia type estimation device 1 can acquire the target signal waveform from the storage device.

[0026] The arrhythmia type estimation device 1 may be installed in a facility where the image analysis device 3 is installed (for example, a medical facility), or it may be installed in a remote location. If installed in a remote location, the arrhythmia type estimation device 1 can acquire the target signal waveform generated by the image analysis device 3 via communication over a communication network such as the Internet.

[0027] The arrhythmia type estimation device 1 may be configured to have a part of the processing performed by another computer. In other words, the processing performed by the arrhythmia type estimation device 1 may be performed by one or more information processing devices. Alternatively, the arrhythmia type estimation device 1 may also have the functions of the image analysis device 3. For example, if the arrhythmia type estimation device 1 also has the functions of the image analysis device 3, the image analysis device 3 is omitted from the components of the arrhythmia type estimation system 100.

[0028] [Display device 4] The display device 4 is a device capable of displaying various information output from the arrhythmia type estimation device 1. In addition to the various information output from the arrhythmia type estimation device 1, the display device 4 may also display input images output from the imaging device 2, target signal waveforms generated by the image analysis device 3, etc.

[0029] (Configuration of arrhythmia type estimation device 1) Next, the configuration of the arrhythmia type estimation device 1 according to one embodiment of the present invention will be described with reference to Figure 3. Figure 3 is a block diagram showing an example of the main components of the arrhythmia type estimation device 1. In the following description, the arrhythmia type estimation device 1, which has the function of estimating the type of tachyarrhythmia that a subject is suffering from from a target signal waveform acquired from an image analysis device 3, will be used as an example.

[0030] As shown in the figure, the arrhythmia type estimation device 1 comprises a processor 10, memory 11, and storage device 12. The arrhythmia type estimation device 1 may be a personal computer, server, or workstation. The processor 10 functions as each of the parts from the acquisition unit 101 to the output control unit 106, which will be described later, by loading the arrhythmia type estimation program 121 stored in the storage device 12 into the memory 11 and executing it.

[0031] The processor 10 can be implemented by logic circuits (hardware) formed on an integrated circuit (IC chip), or by software. When implemented by software, the processor 10 may be composed of, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a combination of these. In this case, the software is stored in the storage device 12. The processor 10 then loads the software into memory 11 and executes it.

[0032] Memory 11 and storage device 12 are both storage devices that store various data used by the arrhythmia type estimation device 1. Memory 11 is a storage device that can write and read data at a higher speed than storage device 12. Storage device 12 has a larger data storage capacity than memory 11. For memory 11, a high-speed access memory such as SDRAM (Synchronous Dynamic Random-Access Memory) can be applied. For storage device 12, for example, an HDD (Hard Disk Drive), SSD (Solid-State Drive), SD (Secure Digital) card, or eMMC (embedded Multi-Media Controller) can be applied.

[0033] Furthermore, the arrhythmia type estimation device 1 is equipped with an input IF unit 13 and an output IF unit 14 as interfaces (IFs) with external devices. The input IF unit 13 is an interface for receiving input signals from input devices such as keyboards and mice, and for acquiring various information and data from external devices. The input IF unit 13 is an interface for, for example, connecting an image analysis device 3 to the input IF unit 13 and acquiring target signal waveforms, etc., from the image analysis device 3. The output IF unit 14 is an interface for outputting the estimation results of the type of tachyarrhythmia, etc., to an external device. The output IF unit 14 can also be connected to a display device, for example, and the estimation results of the type of tachyarrhythmia, etc., can be displayed on the display device.

[0034] The processor 10 functions as an acquisition unit 101, an estimation unit 105, and an output control unit 106, respectively, by executing an arrhythmia type estimation program 121 to acquire target signal waveforms from the image analysis device 3.

[0035] The processor 10 may further include a selection unit (not shown) that selects target signal waveforms from the target signal waveforms acquired from the image analysis device 3 that have a frequency higher than a reference frequency band predetermined as the normal range for pulse rate. In this case, the acquisition unit 101 acquires the target signal waveform selected by the selection unit. Alternatively, the processor 10 may not have a selection unit, but rather the acquisition unit 101 (or estimation unit 105) may have the function of the selection unit.

[0036] The estimation unit 105 estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform acquired by the acquisition unit 101.

[0037] The estimation unit 105 may be configured to estimate the type of tachyarrhythmia the subject is suffering from from the target signal waveform using a trained estimation model 1051. For example, the estimation model 1051 may be machine-trained using training data in which the input waveform generated using sample signal waveforms generated by analyzing cardiac images of multiple sample subjects suffering from tachyarrhythmia is used as the explanatory variable, and the type information indicating the type of tachyarrhythmia each of the multiple sample subjects is used as the objective variable. Here, the sample subjects only need to be of the same biological species as the target subject. The sample signal waveform shows the time-series change in the area of ​​the cardiac region of a sample subject whose type of tachyarrhythmia has been previously identified (diagnosed) by a medical professional such as a physician. The type information is the type of tachyarrhythmia that a medical professional has previously identified for each sample subject.

[0038] The shape of the sample signal waveform exhibits characteristics corresponding to the type of tachyarrhythmia the subject is suffering from. According to the above configuration, a machine learning-based estimation model 1051 can be generated that shows the relationship between the type of tachyarrhythmia occurring in the subject's heart and the shape of the sample signal waveform generated using the sample signal waveform generated from the image of the heart. Using the trained estimation model 1051, the type of tachyarrhythmia the subject is suffering from can be accurately estimated from the shape of the target signal waveform based on the image of the subject's heart.

[0039] The process by which the estimation unit 105 estimates the type of tachyarrhythmia the subject is suffering from based on the target signal waveform, using the estimation model 1051, will be explained with reference to Figure 4. Figure 4 is a diagram illustrating an example of the process performed by the estimation unit 105.

[0040] The target signal waveform RA1 shown in the upper part of Figure 4 represents the time-series change in the area of ​​the right atrium region of a subject's heart. This target signal waveform RA1 beats 10 times in 3 seconds and has a frequency band (approximately 200 bpm) that is higher than the reference frequency band predetermined as the normal range, indicating that this subject suffers from tachyarrhythmia. When the acquisition unit 101 acquires the target signal waveform RA1, the estimation unit 105 inputs the target signal waveform RA1 to the estimation model 1051, and outputs "AFL," which is the type of tachyarrhythmia the subject suffers from, from the estimation model 1051.

[0041] On the other hand, the target signal waveform RA2 shown in the lower part of Figure 4 shows the time-series change in the area of ​​the right atrium region of the heart of another subject. This target signal waveform RA2 beats 9 times in 3 seconds and has a frequency band (approximately 180 bpm) that is higher than the reference frequency band predetermined as the normal range, indicating that this subject suffers from tachyarrhythmia. When the acquisition unit 101 acquires the target signal waveform RA2, the estimation unit 105 inputs the target signal waveform RA2 to the estimation model 1051, and outputs "SVT," which is the type of tachyarrhythmia the subject suffers from.

[0042] The estimation unit 105 may be configured to extract waveforms corresponding to one or more heartbeats of the subject from the target signal waveforms RA1 and RA2, and input the extracted waveforms to the estimation model 1051. Alternatively, the estimation unit 105 may be configured to extract waveforms for a predetermined time from the target signal waveforms RA1 and RA2, and input the extracted waveforms to the estimation model 1051. In the example shown in Figure 4, a 3-second waveform extracted from the target signal waveforms RA1 and RA2 is input to the estimation model 1051.

[0043] Figure 4 shows a case where target signal waveforms RA1 and RA2, which represent the time-series change in the area of ​​the right atrium region of the subject in the input image, are used, but the configuration is not limited to this. That is, the estimation unit 105 may estimate the type of tachyarrhythmia the subject is suffering from using target signal waveforms that represent the time-series change in the area of ​​any of the regions of the right atrium, left atrium, right ventricle, and left ventricle of the subject in the input image.

[0044] Returning to Figure 3, the output control unit 106 causes the estimation results from the estimation unit 105 to be output to various output devices. For example, if a display device 4 is connected via the output IF unit 14, the output control unit 106 may display the estimation results on the display device 4. The manner in which the estimation results are output is arbitrary, and the output control unit 106 may output the estimation results by display output, audio output, print output, or a combination thereof.

[0045] (Processing performed by arrhythmia type estimation device 1) Next, the processing performed by the arrhythmia type estimation device 1 (arrhythmia type estimation method) will be explained using Figure 5. Figure 5 is a flowchart showing an example of the processing flow performed by the arrhythmia type estimation device 1.

[0046] First, the acquisition unit 101 acquires the target signal waveform (step S1: acquisition step). Next, if the target signal waveform acquired by the acquisition unit 101 has a frequency band higher than the reference frequency band predetermined as the normal range (YES in step S2), the estimation unit 105 estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform (step S3: estimation step). Then, the output control unit 106 outputs the estimation result from the estimation unit 105 to an output device (for example, a display device 4) (step S4: output step).

[0047] On the other hand, if the target signal waveform acquired by the acquisition unit 101 does not have frequencies in a frequency band higher than the reference frequency band (NO in step S2), the estimation unit 105 does not estimate the type of tachyarrhythmia.

[0048] With the above configuration, the arrhythmia type estimation device 1 can accurately estimate the type of tachyarrhythmia the subject is suffering from from images of the subject's heart. For example, the arrhythmia type estimation device 1 can estimate with high accuracy whether the fetal tachyarrhythmia is "AFL" or "SVT".

[0049] [Embodiment 2] Other embodiments of the present invention are described below. For the sake of clarity, components having the same function as those described in the above embodiments will be denoted by the same reference numerals, and their descriptions will not be repeated.

[0050] The arrhythmia type estimation device 1 according to the above embodiment is configured to estimate the type of tachyarrhythmia the subject is suffering from by using the target signal waveform as is. On the other hand, the arrhythmia type estimation device 1a according to this embodiment is configured to estimate the type of tachyarrhythmia the subject is suffering from by using one or more target single-beat waveforms corresponding to each heartbeat of the subject's heart from the target signal waveform.

[0051] Similar to the embodiment described above, the following description will use an arrhythmia type estimation device 1a as an example, which has the function of estimating the type of tachyarrhythmia the subject is suffering from from a target signal waveform acquired from an image analysis device 3. However, the arrhythmia type estimation device 1a may also have the functions of the image analysis device 3. In this case, the arrhythmia type estimation device 1a can generate a target signal waveform from an input image acquired from an imaging device 2 and estimate the type of tachyarrhythmia the subject is suffering from.

[0052] (Configuration of arrhythmia type estimation device 1a) The configuration of the arrhythmia type estimation device 1a according to one embodiment of the present invention will be described with reference to Figure 6. Figure 6 is a block diagram showing an example of the main components of the arrhythmia type estimation device 1a.

[0053] As shown in the figure, the arrhythmia type estimation device 1a comprises a processor 10a, memory 11, and storage device 12. The arrhythmia type estimation device 1a may be a personal computer, server, or workstation. The processor 10a functions as each of the parts from the acquisition unit 101 to the output control unit 106, which will be described later, by loading the arrhythmia type estimation program 121 stored in the storage device 12 into the memory 11 and executing it.

[0054] The processor 10a can be implemented by logic circuits (hardware) formed on an integrated circuit (IC chip), or by software. When implemented by software, the processor 10a may be configured as a CPU, a GPU, or a combination of these. In this case, the software is stored in the storage device 12. The processor 10a then loads the software into memory 11 and executes it.

[0055] The processor 10a functions as an acquisition unit 101 that acquires target signal waveforms from the image analysis device 3, a single-beat waveform acquisition unit 102, a first preprocessing unit 103, an estimation unit 105a, and an output control unit 106, by executing the arrhythmia type estimation program 121.

[0056] The beat waveform acquisition unit 102 acquires one or more target beat waveforms corresponding to each heartbeat of the target subject's heart from the target signal waveform acquired by the acquisition unit 101.

[0057] Here, the process by which the single-beat waveform acquisition unit 102 acquires one or more target single-beat waveforms from the target signal waveform will be explained using Figures 7 and 8. Figures 7 and 8 are diagrams illustrating examples of the processes performed by the single-beat waveform acquisition unit 102.

[0058] When the acquisition unit 101 acquires the target signal waveform RA1, the single-beat waveform acquisition unit 102 first analyzes the pattern of area changes associated with the heartbeat of the subject in the target signal waveform RA1 and detects one or more points in time that mark the boundary of each heartbeat of the subject. For example, the diagram shown in the upper right of Figure 7 shows how the single-beat waveform acquisition unit 102 detects the point in time when the heart transitions from systole to diastole (indicated by the star in the diagram) as the boundary of each heartbeat of the subject from the target signal waveform RA1. Subsequently, the single-beat waveform acquisition unit 102 acquires target single-beat waveforms RA1-1 to RA1-10, which are extracted from the target signal waveform RA1 based on the detected boundaries. Figure 7 shows each of the 10 target single-beat waveforms acquired from the target signal waveform RA1.

[0059] When the acquisition unit 101 acquires the target signal waveform RA2, the single-beat waveform acquisition unit 102 first analyzes the pattern of area changes associated with the heartbeat of the subject in the target signal waveform RA2 and detects one or more points in time that mark the boundary of each heartbeat of the subject. For example, the diagram shown in the upper right of Figure 8 shows how the single-beat waveform acquisition unit 102 detects the point in time when the heart transitions from systole to diastole (indicated by the star in the diagram) as the boundary of each heartbeat of the subject from the target signal waveform RA2. Subsequently, the single-beat waveform acquisition unit 102 acquires target single-beat waveforms RA2-1 to RA2-8, which are extracted from the target signal waveform RA2 based on the detected boundary points. Figure 8 shows each of the eight target single-beat waveforms acquired from the target signal waveform RA2.

[0060] Figures 7 and 8 show an example in which the single-beat waveform acquisition unit 102 detects the point in time when the heart transitions from systole to diastole (indicated by the star in the figure) in order to acquire the target single-beat waveform. However, the single-beat waveform acquisition unit 102 may also be capable of detecting any point in time that marks the boundary between each beat of the target subject's heart in the target signal waveform in order to acquire the target single-beat waveform.

[0061] For example, the single-beat waveform acquisition unit 102 may be configured to detect the point in time when the area of ​​the cardiac region in the target signal waveform reaches a predetermined value (e.g., 14,000 pixels) during diastole. In this case, the single-beat waveform acquisition unit 102 acquires the waveform from the point in time when the area of ​​the cardiac region reaches a predetermined value during diastole to the point in time when the area of ​​the cardiac region reaches a predetermined value during the next diastole as the target single-beat waveform.

[0062] The beat waveform acquisition unit 102 may use a learning model such as a convolutional neural network to detect the point in time that marks the boundary between each heartbeat of the subject from the target signal waveform. Such a learning model can be constructed, for example, by machine learning using training data in which the sample signal waveform of the sample subject is used as the explanatory variable and the point in time that marks the boundary between each heartbeat in the said sample signal waveform is used as the objective variable.

[0063] Furthermore, while Figures 7 and 8 show a case where the single-beat waveform acquisition unit 102 acquires a target single-beat waveform from target signal waveforms RA1 and RA2, which show the time-series change in the area of ​​the right atrium region of the subject in the input image, the configuration is not limited to this. That is, the single-beat waveform acquisition unit 102 can acquire a target single-beat waveform from a target signal waveform that shows the time-series change in the area of ​​any of the right atrium, left atrium, right ventricle, and left ventricle regions of the subject in the input image.

[0064] Returning to Figure 6, the first preprocessing unit 103 performs a first preprocessing operation that includes at least one of the following: standardization of the amplitude, which indicates the magnitude of the change in the area of ​​the cardiac region, and standardization of the duration of one heartbeat, for each of the one or more target single-beat waveforms. Here, amplitude standardization is a process of calculating the mean and standard deviation of the area data group for each of the target single-beat waveforms, and dividing the difference between the numerical value of each area data group and the mean by the standard deviation. Through this standardization, the area data group for each of the target single-beat waveforms will have a mean of 0 and a standard deviation of 1, thereby eliminating the effects of differences in the area of ​​the region in the target single-beat waveform (i.e., differences in the size of the heart) and differences in the magnitude of the change in area (amplitude). On the other hand, standardization of the duration of one heartbeat is a process of making the width in the time axis direction of each of the target single-beat waveforms the same.

[0065] Because there are individual differences in heart size and heartbeat duration, the area and duration of the target single-beat waveform obtained from the target signal waveform differ for each subject. Therefore, the first preprocessing unit 103 performs the first preprocessing on each of the one or more target single-beat waveforms.

[0066] The first preprocessing performed by the first preprocessing unit 103 will be explained using Figure 9. Figure 9 is a diagram illustrating an example of the first preprocessing. Figure 9 shows an example where the first preprocessing unit 103 performs amplitude standardization and standardization of the duration of one heartbeat.

[0067] The upper part of Figure 9 shows the target beat waveform RA1-xN after the first preprocessing has been performed on each of the multiple target beat waveforms RA1-x obtained from the target signal waveform RA1. On the other hand, the lower part of Figure 9 shows the target beat waveform RA2-xN after the first preprocessing has been performed on each of the multiple target beat waveforms RA2-x obtained from the target signal waveform RA2. As shown in Figure 9, the shape of the target beat waveform RA1-xN has the same characteristics as the shape of the target beat waveform RA1-x before the first preprocessing is performed, and the shape of the target beat waveform RA2-xN has the same characteristics as the shape of the target beat waveform RA2-x before the first preprocessing is performed.

[0068] Returning to Figure 6, the estimation unit 105a estimates the type of tachyarrhythmia the subject is suffering from based on the shape of one or more target single-beat waveforms after the first preprocessing. Here, the estimation unit 105a may be configured to estimate the type of tachyarrhythmia the subject is suffering from from one or more target single-beat waveforms after the first preprocessing using a trained estimation model 1051a. In this case, the input waveforms used as explanatory variables in machine learning to generate the estimation model 1051a may be generated by performing a third preprocessing on each of the one or more sample single-beat waveforms corresponding to each heartbeat of the subject's heart in the sample signal waveform. The third preprocessing is the same process as the first preprocessing performed on each of the one or more target sample single-beat waveforms. Specifically, the third preprocessing is a process that includes, for each of the one or more sample single-beat waveforms, at least one of the following: standardization of the amplitude indicating the magnitude of the area change of the cardiac region, and standardization of the length of time of one heartbeat.

[0069] With the above configuration, a machine learning-based estimation model 1051a can be generated that recognizes the relationship between the type of tachyarrhythmia occurring in the heart of a sample subject and the shape of a sample single-beat waveform generated using a sample signal waveform generated from an image of the heart. Using the trained estimation model 1051a, the type of tachyarrhythmia the subject is suffering from can be accurately estimated from the shape of a target single-beat waveform generated from a target signal waveform based on an image of the subject's heart.

[0070] The process by which the estimation unit 105a estimates the type of tachyarrhythmia the subject is suffering from, based on the waveform of a single beat after the first preprocessing, using the estimation model 1051a, will be explained with reference to Figure 10. Figure 10 is a diagram illustrating an example of the process performed by the estimation unit 105a.

[0071] The upper part of Figure 10 shows the target single-beat waveform RA1-N after first preprocessing, obtained from target signal waveforms RA1 and RA2, which show the time-series change in the area of ​​the right atrium region of the heart of a subject suffering from tachyarrhythmia. The estimation unit 105a inputs the target single-beat waveform RA1-N after the first preprocessing to the estimation model 1051a, and outputs "AFL," which is the type of tachyarrhythmia the subject is suffering from, from the estimation model 1051a.

[0072] Meanwhile, the lower part of Figure 10 shows the target single-beat waveform RA2-N after performing the first preprocessing, which was obtained from the target signal waveform RA2, showing the time-series change in the area of ​​the right atrium region of the heart of another subject suffering from tachyarrhythmia. The estimation unit 105a inputs the target single-beat waveform RA2-N after performing the first preprocessing to the estimation model 1051a, and outputs "SVT," which is the type of tachyarrhythmia the subject is suffering from, from the estimation model 1051a.

[0073] Figure 10 shows a configuration in which one target single-beat waveform is input to the estimation model 1051a and one estimation result is output, but the estimation unit 105a is not limited to this configuration. The estimation unit 105a may be configured to estimate the type of tachyarrhythmia the subject is suffering from, based on one or more estimation results output from the estimation model 1051a, each of which has been input to one or more target single-beat waveforms after the first preprocessing.

[0074] For example, the estimation unit 105a may input each of the six target beat waveforms shown in Figure 9 (RA1-xN) to the estimation model 1051a. In this case, if a predetermined percentage (e.g., 60%) or more of the six estimation results output from the estimation model 1051a are "AFL", the estimation unit 105a may estimate that the type of tachyarrhythmia the subject is suffering from is "AFL". Similarly, for example, the estimation unit 105a may input each of the eight target beat waveforms shown in Figure 9 (RA2-xN) to the estimation model 1051a. In this case, if a predetermined percentage (e.g., 60%) or more of the eight estimation results output from the estimation model 1051a are "SVT", the estimation unit 105a may estimate that the type of tachyarrhythmia the subject is suffering from is "SVT".

[0075] (Processing performed by arrhythmia type estimation device 1a) Next, the processing performed by the arrhythmia type estimation device 1a (arrhythmia type estimation method) will be explained using Figure 11. Figure 11 is a flowchart showing an example of the processing flow performed by the arrhythmia type estimation device 1a.

[0076] First, the acquisition unit 101 acquires the target signal waveform (step S1: acquisition step). Next, if the target signal waveform acquired by the acquisition unit 101 has a frequency in a frequency band higher than the reference frequency band predetermined as the normal range (YES in step S2), the single beat waveform acquisition unit 102 acquires one or more target single beat waveforms from the target signal waveform (step S3a: single beat waveform acquisition step).

[0077] Next, the first preprocessing unit 103 performs a first preprocessing step for each of the one or more target beat waveforms that have been acquired (step S3b: first preprocessing step).

[0078] Next, the estimation unit 105a estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target single-beat waveform after the first preprocessing (step S3c: estimation step). Then, the output control unit 106 outputs the estimation result from the estimation unit 105a to an output device (for example, a display device 4) (step S4: output step).

[0079] With the above configuration, the arrhythmia type estimation device 1a can more accurately estimate the type of tachyarrhythmia the subject is suffering from from an image of the subject's heart. [Embodiment 3] Other embodiments of the present invention are described below. For the sake of clarity, components having the same function as those described in the above embodiments will be denoted by the same reference numerals, and their descriptions will not be repeated.

[0080] Similar to the embodiment described above, the following description will use an arrhythmia type estimation device 1b as an example, which has the function of estimating the type of tachyarrhythmia the subject is suffering from from a target signal waveform acquired from an image analysis device 3. However, the arrhythmia type estimation device 1b may also have the functions of the image analysis device 3. In this case, the arrhythmia type estimation device 1b generates a target signal waveform from an input image acquired from an imaging device 2 and estimates the type of tachyarrhythmia the subject is suffering from.

[0081] (Configuration of arrhythmia type estimation device 1b) The configuration of the arrhythmia type estimation device 1b according to one embodiment of the present invention will be described with reference to Figure 12. Figure 12 is a block diagram showing an example of the main components of the arrhythmia type estimation device 1b.

[0082] As shown in the figure, the arrhythmia type estimation device 1b comprises a processor 10b, memory 11, and storage device 12. The arrhythmia type estimation device 1b may be a personal computer, server, or workstation. The processor 10b functions as each of the parts from the acquisition unit 101 to the output control unit 106, which will be described later, by loading the arrhythmia type estimation program 121 stored in the storage device 12 into the memory 11 and executing it.

[0083] The processor 10b can be implemented by logic circuits (hardware) formed on an integrated circuit (IC chip), or by software. When implemented by software, the processor 10b may be configured as a CPU, a GPU, or a combination of these. In this case, the software is stored in the storage device 12. The processor 10b then loads the software into memory 11 and executes it.

[0084] The processor 10b functions as an acquisition unit 101 that acquires target signal waveforms from the image analysis device 3, a single-beat waveform acquisition unit 102, a first preprocessing unit 103, a second preprocessing unit 104, an estimation unit 105b, and an output control unit 106 by executing the arrhythmia type estimation program 121.

[0085] The second preprocessing unit 104 performs a second preprocessing step for each of the one or more target beat waveforms, extracting a predetermined period of interest from the target beat waveform as a target partial waveform. In other words, the second preprocessing unit 104 generates a target partial waveform by performing the first preprocessing step and the second preprocessing step for each of the one or more target beat waveforms.

[0086] The estimation unit 105b estimates the type of tachyarrhythmia the subject is suffering from based on the shape of one or more target partial waveforms. Here, the estimation unit 105b may be configured to estimate the type of tachyarrhythmia the subject is suffering from from one or more target partial waveforms using a trained estimation model 1051b. In this case, the input waveforms used as explanatory variables in machine learning to generate the estimation model 1051b may be generated by performing third and fourth preprocessing on each of the one or more sample beat waveforms that correspond to each heartbeat of the subject's heart in the sample signal waveform. The fourth preprocessing is the process of extracting a predetermined period of interest from each of the one or more target beat waveforms.

[0087] For example, when the type of tachyarrhythmia is "AFL," the shape of the target beat waveform RA1-N in the period where the horizontal axis representing normalized time is greater than or equal to 0.6 to 0.8 is characteristic and differs significantly from the case where the type of tachyarrhythmia is "SVT." On the other hand, the difference in the target beat waveform RA1-N in the period corresponding to 0.6 to 0.8 or greater on the horizontal axis representing normalized time is small compared to the case where the type of tachyarrhythmia is "SVT." Therefore, it is possible to extract a target partial waveform from the target beat waveform that corresponds to the period of interest having a characteristic shape for each type of tachyarrhythmia, and estimate the type of tachyarrhythmia using the target partial waveform.

[0088] The process by which the second preprocessing unit 104 extracts a target partial waveform from the target single-beat waveform, and the process by which the estimation unit 105b estimates the type of tachyarrhythmia the subject is suffering from from the target partial waveform, will be explained using Figure 13. Figure 13 is a diagram illustrating an example of the processes performed by the second preprocessing unit 104 and the estimation unit 105b.

[0089] The target single-beat waveform RA1-N shown in the upper part of Figure 13 represents the time-series change in the area of ​​the right atrium region of a subject's heart. The second preprocessing unit 104 extracts a portion of the target single-beat waveform RA1-N that corresponds to a predetermined period of interest P1. The estimation unit 105b inputs the target portion waveform RA1-P1 extracted by the second preprocessing unit 104 into the estimation model 1051b and estimates that the type of tachyarrhythmia the subject is suffering from is "AFL".

[0090] The target single-beat waveform RA2-N shown in the lower part of Figure 13 represents the time-series change in the area of ​​the right atrium region of the heart of another subject. The second preprocessing unit 104 extracts a predetermined period of interest P2 from the target single-beat waveform RA2-N. The estimation unit 105b inputs the target partial waveform RA2-P2 extracted by the second preprocessing unit 104 into the estimation model 1051b and estimates that the type of tachyarrhythmia the subject is suffering from is "SVT".

[0091] Figure 13 shows an example where the target partial waveforms RA1-P1 and RA2-P2 are used in focus periods P1 and P2, which correspond to a range of 0 to 0.75 on the horizontal axis representing normalized time. However, the configuration is not limited to this, and focus periods P1 and P2 can be set to any period in which the differences in waveforms due to the type of tachyarrhythmia the subject suffers from are significant.

[0092] Figure 13 shows a configuration in which one target beat waveform is input to the estimation model 1051b and one estimation result is output, but the estimation unit 105b is not limited to this configuration. The estimation unit 105b may be configured to estimate the type of tachyarrhythmia the subject is suffering from, based on one or more estimation results output from the estimation model 1051b, each of which has been input to one or more target partial waveforms after first preprocessing and second preprocessing of the target signal waveform.

[0093] For example, the estimation unit 105b may estimate that the type of tachyarrhythmia the subject has is "AFL" if a predetermined proportion (e.g., 60%) or more of the one or more estimation results output from the estimation model 1051b are "AFL". Alternatively, the estimation unit 105b may estimate that the type of tachyarrhythmia the subject has is "SVT" if a predetermined proportion (e.g., 60%) or more of the one or more estimation results output from the estimation model 1051b are "SVT".

[0094] (Processing performed by arrhythmia type estimation device 1b) Next, the processing performed by the arrhythmia type estimation device 1b (arrhythmia type estimation method) will be explained using Figure 14. Figure 14 is a flowchart showing an example of the processing flow performed by the arrhythmia type estimation device 1b.

[0095] First, the acquisition unit 101 acquires the target signal waveform (step S1: acquisition step). Next, if the target signal waveform acquired by the acquisition unit 101 has a frequency in a frequency band higher than the reference frequency band predetermined as the normal range (YES in step S2), the single beat waveform acquisition unit 102 acquires one or more target single beat waveforms from the target signal waveform (step S3a: single beat waveform acquisition step).

[0096] Next, the first preprocessing unit 103 performs a first preprocessing step on each of the acquired target beat waveforms (step S3b: first preprocessing step), and the second preprocessing unit 104 extracts a target partial waveform from each of the target beat waveforms after the first preprocessing step has been performed (step S3c: second preprocessing step).

[0097] Next, the estimation unit 105b estimates the type of tachyarrhythmia the subject is suffering from based on the shape of one or more target partial waveforms (step S3d: estimation step). Then, the output control unit 106 outputs the estimation result from the estimation unit 105b to the output device (step S4: output step).

[0098] On the other hand, if the target signal waveform acquired by the acquisition unit 101 does not have frequencies in a frequency band higher than the reference frequency band (NO in step S2), the estimation unit 105b does not estimate the type of tachyarrhythmia.

[0099] According to the above configuration, the arrhythmia type estimation device 1b extracts a target partial waveform containing a characteristic shape corresponding to the type of tachyarrhythmia from each of the one or more target single-beat waveforms after the first preprocessing, and estimates the type of tachyarrhythmia the subject is suffering from using the target partial waveform. As a result, the arrhythmia type estimation device 1b can estimate the type of tachyarrhythmia the subject is suffering from with even greater accuracy.

[0100] [Examples of implementation using software] The functions of the arrhythmia type estimation devices 1, 1a, and 1b (hereinafter referred to as "devices") are programs that cause a computer to function as the device, and these programs can be realized by programs that cause a computer to function as each control block of the device (particularly each part included in processors 10, 10a, and 10b).

[0101] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, the functions described in each of the embodiments are realized.

[0102] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0103] Furthermore, some or all of the functions of each of the above control blocks can also be realized by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to realize the functions of each of the above control blocks by, for example, a quantum computer.

[0104] Furthermore, each process described in the above embodiments may be performed by AI (Artificial Intelligence). In this case, the AI ​​may operate on the control device described above, or it may operate on other devices (for example, an edge computer or a cloud server).

[0105] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

[0106] 〔summary〕 An arrhythmia type estimation device according to embodiment 1 of the present invention comprises: an acquisition unit that acquires a target signal waveform showing the time-series change in the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing an image of the heart of a subject; and an estimation unit that estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as the normal range.

[0107] An arrhythmia type estimation device according to embodiment 2 of the present invention further comprises, in embodiment 1 above, a beat waveform acquisition unit that acquires one or more target beat waveforms corresponding to each of the heartbeats of the subject from the target signal waveform, and a first preprocessing unit that performs a first preprocessing including at least one of the following: standardization of the amplitude indicating the magnitude of the area change of the region shown in each of the one or more target beat waveforms, and standardization of the length of time of one of the heartbeats of the heart, wherein the estimation unit may estimate the type of tachyarrhythmia that the subject is suffering from based on the shape of the one or more target beat waveforms after the first preprocessing.

[0108] The arrhythmia type estimation device according to embodiment 3 of the present invention further comprises a second preprocessing unit which performs a second preprocessing for each of the one or more target beat waveforms, in embodiment 1 or 2 above, to extract a predetermined period of interest portion of the target beat waveform as a target partial waveform, and the estimation unit may estimate the type of tachyarrhythmia that the target subject is suffering from based on the shape of one or more of the target partial waveforms.

[0109] The arrhythmia type estimation device according to aspect 4 of the present invention may be configured such that, in any of aspects 1 to 3 above, the estimation unit uses an input waveform generated using a sample signal waveform that shows the time-series change of the area of ​​the region, which is generated by analyzing cardiac images of multiple sample subjects suffering from tachyarrhythmias, as an explanatory variable, and uses a machine learning estimation model that uses training data in which type information indicating the type of tachyarrhythmias each of the multiple sample subjects suffers from is used as the objective variable, to estimate the type of tachyarrhythmias that the target subject suffers from from the target signal waveform.

[0110] The arrhythmia type estimation device according to aspect 5 of the present invention, in aspect 4 above, is configured such that the input waveform is generated by performing a third preprocessing on each of the one or more sample beat waveforms corresponding to each heartbeat of the subject in the sample signal waveform, which includes at least one of the following: standardization of the amplitude indicating the magnitude of the area change in the region, and standardization of the length of time of each heartbeat of the heart, and the estimation unit uses the estimation model to estimate the type of tachyarrhythmia that the subject is suffering from from the one or more target beat waveforms corresponding to each heartbeat of the subject obtained from the target signal waveform, which have been subjected to the same preprocessing as the third preprocessing.

[0111] The arrhythmia type estimation device according to embodiment 6 of the present invention may be configured such that, in embodiment 5, the estimation unit estimates the type of tachyarrhythmia the subject is suffering from based on one or more estimation results output from the estimation model, which is input to each of the one or more target single-beat waveforms after performing the same preprocessing as the third preprocessing on the target signal waveform.

[0112] The arrhythmia type estimation device according to embodiment 7 of the present invention, in embodiment 4 above, is configured to generate the input waveform by performing a third preprocessing step for each of the one or more sample beat waveforms corresponding to each heartbeat of the sample subject's heart in the sample signal waveform, which includes at least one of the following: standardization of the amplitude indicating the magnitude of the area change of the region and standardization of the length of time of one heartbeat of the heart; and a fourth preprocessing step for extracting a predetermined period of interest from each of the one or more target beat waveforms of the sample beat waveform. The estimation unit may be configured to estimate the type of tachyarrhythmia the subject is suffering from from one or more target partial waveforms generated by performing a first preprocessing step for each of the one or more target beat waveforms, which includes at least one of the following: standardization of the amplitude indicating the magnitude of the area change of the region shown in each of the one or more target beat waveforms and standardization of the length of time of one heartbeat of the heart; and a second preprocessing step for each of the one or more target beat waveforms, which extracts a predetermined period of interest from the target beat waveform as a target partial waveform, using the estimation model.

[0113] The arrhythmia type estimation device according to embodiment 8 of the present invention may be configured such that, in embodiment 7, the estimation unit estimates the type of tachyarrhythmia that the subject is suffering from based on one or more estimation results output from the estimation model, each of which has been input one or more of the target partial waveforms after the first preprocessing and second preprocessing have been performed on the target signal waveform.

[0114] The arrhythmia type estimation device according to embodiment 9 of the present invention is, in any of embodiments 1 to 8 above, wherein the region is at least one of the left atrium and the right atrium, and the estimation unit may estimate whether the type of tachyarrhythmia the subject is suffering from is supraventricular tachycardia or atrial flutter based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as the normal range.

[0115] An arrhythmia type estimation method according to embodiment 10 of the present invention is an arrhythmia type estimation method performed by one or more information processing devices, comprising: an acquisition step of acquiring a target signal waveform that shows the time-series change of the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing an image of the heart of a subject; and an estimation step of estimating the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency in a frequency band higher than a reference frequency band predetermined as a normal range.

[0116] The arrhythmia type estimation program according to aspect 11 of the present invention is an arrhythmia type estimation program for causing a computer to function as an arrhythmia type estimation device according to any of aspects 1 to 9 above, and is an arrhythmia type estimation program for causing a computer to function as the acquisition unit and the estimation unit. [Explanation of Symbols]

[0117] 1, 1a, 1b Arrhythmia type estimation device 10, 10a, 10b processors 101 Acquisition Department 102 Single beat waveform acquisition section 103 First Pre-processing Unit 104 Second Pre-processing Unit 105, 105a, 105b Estimation section 121 Arrhythmia Type Estimation Program Estimated models 1051, 1051a, and 1051b

Claims

1. An acquisition unit acquires a target signal waveform that shows the time-series change in the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing images of the heart of the subject, An estimation unit that estimates the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency band higher than a reference frequency band predetermined as the normal range, An arrhythmia type estimation device equipped with the following features.

2. A beat waveform acquisition unit acquires one or more target beat waveforms corresponding to each heartbeat of the target subject's heart from the target signal waveform, A first preprocessing unit that performs a first preprocessing including at least one of the following: standardization of the amplitude indicating the magnitude of the area change of the region shown in each of the one or more target single beat waveforms, and standardization of the length of time of one heartbeat of the heart; Furthermore, The estimation unit, Based on the shape of the one or more target single-beat waveforms after the first preprocessing, the type of tachyarrhythmia the target subject is suffering from is estimated. The arrhythmia type estimation device according to claim 1.

3. The system further includes a second preprocessing unit which performs a second preprocessing operation on each of the one or more target beat waveforms, in which it extracts a predetermined period of interest portion of the target beat waveform as a target partial waveform. The estimation unit, Based on the shape of one or more of the target partial waveforms, the type of tachyarrhythmia the subject is suffering from is estimated. The arrhythmia type estimation device according to claim 2.

4. The estimation unit, The method involves using a machine learning-based estimation model, which uses training data with an input waveform generated from sample signal waveforms showing the time-series change in the area of ​​a region, generated by analyzing cardiac images of multiple sample subjects suffering from tachyarrhythmias, as an explanatory variable, and type information indicating the type of tachyarrhythmias each of the multiple sample subjects suffers from as an objective variable, to estimate the type of tachyarrhythmias a subject suffers from from the target signal waveform. The arrhythmia type estimation device according to claim 1.

5. The input waveform is generated by performing a third preprocessing on each of the one or more sample beat waveforms in the sample signal waveform that correspond to each heartbeat of the sample subject's heart, which includes at least one of the following: standardization of the amplitude indicating the magnitude of the area change in the region, and standardization of the duration of each heartbeat of the heart. The estimation unit, Using the estimation model, one or more target beat waveforms corresponding to each heartbeat of the subject obtained from the target signal waveform, after performing the same preprocessing as the third preprocessing, are used to estimate the type of tachyarrhythmia the subject is suffering from. The arrhythmia type estimation device according to claim 4.

6. The estimation unit, Based on one or more estimation results output from the estimation model to which each of the one or more target single-beat waveforms, after performing the same preprocessing as the third preprocessing on the target signal waveform, the type of tachyarrhythmia the subject is suffering from is estimated. The arrhythmia type estimation device according to claim 5.

7. The aforementioned input waveform is A third preprocessing step is performed on each of the one or more sample beat waveforms in the sample signal waveform that correspond to each heartbeat of the sample subject, including at least one of the following: standardization of the amplitude showing the magnitude of the area change in the region, and standardization of the duration of each heartbeat of the heart; A fourth preprocessing step involves extracting a predetermined period of interest from each of the one or more target beat waveforms from the aforementioned sample beat waveform. It is generated by performing the following: The estimation unit, Using the aforementioned estimation model, A first preprocessing step, which includes at least one of the following: standardization of the amplitude indicating the magnitude of the area change in the region shown in each of the one or more target single-beat waveforms, and standardization of the duration of one heartbeat of the heart; A second preprocessing step involves extracting a portion of the aforementioned target beat waveform corresponding to a predetermined period of interest as the target partial waveform. By performing this process for each of the one or more target single-beat waveforms, the type of tachyarrhythmia the subject is suffering from is estimated from the one or more target partial waveforms generated. The arrhythmia type estimation device according to claim 4.

8. The estimation unit, Based on one or more estimation results output from the estimation model, which receives each of the one or more target partial waveforms after performing the first and second preprocessing on the target signal waveform, the type of tachyarrhythmia the subject is suffering from is estimated. The arrhythmia type estimation device according to claim 7.

9. The region is at least one of the left atrium and the right atrium. The estimation unit, Based on the shape of the target signal waveform having frequencies in a frequency band higher than a predetermined reference frequency band defined as the normal range, it is estimated whether the type of tachyarrhythmia the subject is suffering from is supraventricular tachycardia or atrial flutter. The arrhythmia type estimation device according to claim 1.

10. A method for estimating the type of arrhythmia, which is performed by one or more information processing devices, Acquisition step: Obtain a target signal waveform that shows the time-series change in the area of ​​at least one of the regions of the left atrium, left ventricle, right atrium, and right ventricle, generated by analyzing images of the heart of the subject; An estimation step of estimating the type of tachyarrhythmia the subject is suffering from based on the shape of the target signal waveform having a frequency band higher than a reference frequency band predetermined as the normal range, A method for estimating arrhythmia types, including those mentioned above.

11. An arrhythmia type estimation program for causing a computer to function as an arrhythmia type estimation device according to claim 1, wherein the acquisition unit and the estimation unit are the same as the computer.

Images