JP7428055B2 - Diagnostic support system, diagnostic support device and program - Google Patents

Description

The present invention relates to a diagnostic support system, a diagnostic support device, and a program.

In recent years, technology has been proposed that utilizes AI (artificial intelligence) for interpretation and diagnosis of medical images.
For example, Patent Document 1 describes that a desired region (for example, a lung field) in a diagnostic X-ray image is divided into rectangular regions and subjected to lesion detection processing by an AI engine.
Furthermore, Non-Patent Document 1 describes that a chest image obtained by an X-ray examination is imported into AI, the presence or absence of a lesion is determined by AI, and the lesion is shown surrounded by a rectangle.

Japanese Patent Application Publication No. 2019-154943

Xiaosong Wang et al.: “ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases”, IEEE CVPR 2017, pp 2097-2106(2017)

Basically, the object of image interpretation by a doctor is a region of interest, and the doctor performs image interpretation by focusing on the region of interest in a medical image.
The technique disclosed in Patent Document 1 can assist a doctor in interpreting and diagnosing a region of interest. However, if an abnormality such as a lesion exists outside the region of interest that was not intended by the doctor, there is a possibility that the abnormality may be overlooked. On the other hand, in the technique of Non-Patent Document 1, the entire chest image is taken in by AI and lesions are determined, which increases the processing load of AI and requires an AI engine with high processing capacity.

An object of the present invention is to assist in finding abnormalities at unintended locations and to efficiently prevent oversights in image interpretation.

In order to solve the above problem, the diagnosis support system of the invention according to claim 1 includes:
a storage means for storing a predetermined operation performed by a user on the medical information of the target test;
identification means for identifying an area outside the region of interest from at least one medical image of the target examination based on the user's predetermined operation on the medical information of the target examination stored in the storage means;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the specifying means;
Equipped with.

The invention according to claim 2 is the invention according to claim 1,
An output means is provided for outputting a detection result by the detection means.

The invention according to claim 3 is the invention according to claim 1 or 2,
The specifying means specifies a region of interest from at least one medical image of the target test based on the user's predetermined operation on the medical information of the target test stored in the storage unit, and A region outside the specified region of interest is identified as a region outside the region of interest.

The invention according to claim 4 is the invention according to any one of claims 1 to 3,
The specifying means further specifies, with respect to another medical image of the same patient as the medical image, in which a region outside the region of interest has not been specified, the entire other medical image to be outside the region of interest;
The detection means performs abnormality detection on the entire other medical image using artificial intelligence.

The invention according to claim 5 is the invention according to any one of claims 1 to 3,
The specifying means further determines, with respect to another medical image of the same patient as the medical image, in which a region outside the region of interest has not been specified, a region corresponding to the region identified as outside the region of interest in the medical image. Identified as outside the area,
The detection means performs abnormality detection using artificial intelligence only in a region of the other medical image that is outside the region of interest of the other medical image.

The invention according to claim 6 is the invention according to claim 4 or 5,
The other medical image is at least one medical image among medical images acquired with a different modality, medical images acquired at different times, or medical images photographed with a different imaging method.

The invention according to claim 7 is the invention according to any one of claims 1 to 6,
The specifying means specifies the entire medical image as a region outside the region of interest when the region outside the region of interest cannot be specified from the medical image based on the predetermined operation stored in the storage means.

The diagnostic support device of the invention according to claim 8 includes:
identification means for receiving a predetermined operation performed by a user on the medical information of the target examination, and identifying an area outside the region of interest from at least one medical image of the target examination based on the predetermined operation;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the specifying means;
Equipped with.

The program of the invention according to claim 9 is:
computer,
storage means for storing predetermined operations performed by the user on the medical information of the target test;
identification means for identifying an area outside the region of interest from at least one medical image of the target examination based on the user's predetermined operation on the medical information of the target examination stored in the storage means;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the identification means;
function as

According to the present invention, it is possible to assist in finding abnormalities at unintended locations, and it is possible to efficiently prevent oversights in image interpretation.

1 is a diagram illustrating an example of the overall configuration of a diagnostic support system according to the present embodiment. FIG. 2 is a block diagram showing the functional configuration of the server device in FIG. 1. FIG. FIG. 2 is a block diagram showing the functional configuration of the image interpretation terminal in FIG. 1. FIG. 3 is a flowchart showing diagnostic support processing executed by the control unit in FIG. 2. FIG. FIG. 6 is a diagram schematically illustrating a user's predetermined operation and identification of areas outside the region of interest. (a) is a diagram showing an example of a notification informing that the interpretation of images outside the region of interest by AI has been completed, and (b) is a diagram showing where the region outside the region of interest was where an abnormality was detected by AI. It is a figure which shows the example of a display of an image. (a) is a diagram showing an example of a notification of an abnormality detection result outside the region of interest by AI, and (b) is a diagram showing an example of display of the position of the abnormality detected by AI.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. Note that the present invention is not limited to the illustrated example.

(Configuration of diagnosis support system 100)
First, the configuration of an embodiment of the present invention will be described.
FIG. 1 shows the system configuration of a diagnosis support system 100 in this embodiment.
As shown in FIG. 1, the diagnosis support system 100 includes an image generation device 1, a server device 2, and an image interpretation terminal 3. These devices 1 to 3 are connected to each other so as to be able to send and receive data via a communication network N, such as a LAN (Local Area Network), constructed within the medical facility. The communication network N is based on the DICOM (Digital Imaging and Communication in Medicine) standard. Note that the number of each device is not particularly limited.

Each component device 1 to 3 will be explained below.
The image generation device 1 is a modality that photographs a subject part of a human body and generates digital data of the photographed image (medical image), and is, for example, a CR (Computed Radiography) device, an FPD (Flat Panel Detector) device, a CT ( Computed Tomography) equipment, MRI (Magnetic Resonance Imaging) equipment, ultrasound diagnostic equipment, etc. can be applied.

Note that the image generation device 1 is a device that complies with the above-mentioned DICOM standard, and generates medical images in the DICOM file format. A DICOM file consists of an image section and a header section. The actual data of the medical image is written in the image section, and the additional information (DICOM information) of the medical image is written in the header section. The image generation device 1 adds the above-mentioned supplementary information as header information to the medical image obtained by photographing, and transmits the medical image to the server device 2 via the communication network N.

Here, the DICOM information includes patient information, examination information, series information, and image information.
The patient information is various information regarding the patient (subject) of the medical image, and includes items such as patient ID, patient name, date of birth, age, sex, height, and weight.
The examination information is various information regarding the examination, and includes items such as examination ID, examination date, examination time, examination description, number of series, and number of images.
Series information is various information related to the series included in the examination, such as modality, imaging site, imaging method (e.g. imaging direction (slice direction), presence or absence of contrast agent), series description, laterality, series number, image Items such as number of sheets and AE title are included.
Image information includes various information related to medical images, such as image number, patient direction, image type, pixel structure, vertical width, horizontal width, tube voltage, mA, tube current, pixel spacing, slice thickness, slice position, and frame number. , number of frames, TR, TE, WW/WC, and other items are included.

The server device 2 stores the medical images transmitted from the image generation device 1 in an image DB (Data Base) 251, stores the supplementary information in the supplementary information DB 252, and manages the medical images. In addition, the server device 2 displays an examination list screen and a viewer screen on the image interpretation terminal 3 in response to a request from the image interpretation terminal 3, and performs various processes in response to user operations.
Here, the examination list screen is a screen on which an examination list for selecting a target examination for image interpretation is displayed. The viewer screen is a screen for displaying medical images of the target examination to support the doctor's interpretation and diagnosis.

FIG. 2 shows an example of the functional configuration of the server device 2. As shown in FIG.
As shown in FIG. 2, the server device 2 includes a control section 21, an operation section 22, a display section 23, a communication section 24, a storage section 25, and a detection section 26, and each section is connected by a bus 27. .

The control unit 21 includes a CPU (Central Processing Unit) and a RAM (Random Access Memory).
) etc. The CPU of the control unit 21 reads out various programs such as system programs and processing programs stored in the storage unit 25, expands them into the RAM, and executes various processes according to the expanded programs. For example, the control unit 21 cooperates with a program stored in the storage unit 25 to execute diagnostic support processing, which will be described later.

The operation unit 22 includes a keyboard with character input keys, number input keys, various function keys, etc., and a pointing device such as a mouse, and receives push signals from keys pressed on the keyboard and operation signals from the mouse. is output to the control section 21 as an input signal.

The display unit 23 includes a monitor such as a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display), and displays various screens according to instructions from display signals input from the control unit 21.

The communication unit 24 includes a LAN card and the like, and transmits and receives data to and from external devices connected to the communication network N.

The storage unit 25 is composed of, for example, an HDD (Hard Disk Drive), a semiconductor nonvolatile memory, or the like. The storage unit 25 stores various programs as described above. The programs stored in the storage unit 25 include programs for executing diagnostic support processing, which will be described later.

Furthermore, the storage unit 25 is provided with an image DB 251 and an additional information DB 252.
The image DB 251 is a database for storing medical images and interpretation reports.
The supplementary information DB 252 is a database that stores supplementary information regarding the medical images stored in the image DB 251 in a searchable manner.

The detection unit 26 performs abnormality detection processing using AI (artificial intelligence) from a designated area of the medical image instructed by the control unit 21, and outputs detection result information.
In this embodiment, the detection unit 26 learns by machine learning using, for example, a large amount of learning data (medical images showing abnormalities (lesions) and correct labels (pairs of abnormal regions and diagnosis names in the medical images). Using the machine learning model created by this, abnormality detection processing is performed from a designated area of the input medical image, and detection result information is output to the control unit 21 in association with the medical image.
Note that the AI that constitutes the detection unit 26 is not limited to the one described above, and may detect an abnormality using any method.

The image interpretation terminal 3 is a terminal equipped with an image interpretation application. The image interpretation application accesses the server device 2 according to the user's operation, displays the examination list screen, viewer screen, etc. sent from the server device 2, accepts the user's operation from these screens, and displays the image according to the user's operation. This is an application that executes processing. As the image interpretation terminal 3, a personal computer, a tablet, a smartphone, etc. can be used.

FIG. 3 shows an example of the functional configuration of the image interpretation terminal 3.
As shown in FIG. 3, the image interpretation terminal 3 includes a control section 31, an operation section 32, a display section 33, a communication section 34, and a storage section 35, and each section is connected by a bus 36.

The control unit 31 is composed of a CPU, RAM, etc. The CPU of the control unit 31 reads out various programs such as a system program and an image interpretation application stored in the storage unit 35, develops them in the RAM, and executes various processes according to the developed programs.

The operation unit 32 includes a keyboard with character input keys, number input keys, various function keys, etc., and a pointing device such as a mouse, and receives push signals from keys pressed by the user and operation signals from the mouse. is output to the control section 31 as an input signal. Further, the operation section 32 may be configured to include a touch panel on the display screen of the display section 33, and in this case, outputs an instruction signal input by user operation via the touch panel to the control section 31.

The display unit 33 includes a monitor such as a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display), and displays various screens according to instructions from display signals input from the control unit 31.

The communication unit 34 is configured with a LAN card or the like, and transmits and receives data to and from an external device connected to the communication network N by wire or wirelessly.

The storage unit 35 is composed of, for example, an HDD (Hard Disk Drive), a semiconductor nonvolatile memory, or the like. The storage unit 35 stores various programs including the system program and the image interpretation application as described above.

(Operation of diagnosis support system 100)
Next, the operation of the diagnosis support system 100 will be explained.
When the image interpretation application is started on the image interpretation terminal 3 and a login request is made to the server device 2, the server device 2 performs login authentication. After the authentication, the server device 2 transmits the examination list screen to the image interpretation terminal 3 through the communication unit 24. When a target examination for image interpretation is selected from the image interpretation terminal 3, the server device 2 generates screen information of the viewer screen on which the medical image of the selected target examination is displayed, and transmits it to the image interpretation terminal 3 via the communication unit 24. , starts diagnostic support processing.

In the image interpretation terminal 3, when the viewer screen is transmitted from the server device 2, the viewer screen is displayed on the display unit 33. The viewer screen can accept various operations by the user (e.g., enlarging/reducing the medical image, panning (moving the display area in the medical image), ROI setting, gradation adjustment, comparative interpretation, creating an interpretation report, exit instructions, etc.) It is configured. The control unit 31 executes processing according to user operations from the operation unit 32 according to the image interpretation application, and transmits operation information of the user operations to the server device 2 via the communication unit 34.

FIG. 4 shows a flowchart of the diagnostic support process executed by the control unit 21 of the server device 2. The diagnosis support process is executed in cooperation with the control unit 21 and a program stored in the storage unit 25. Further, the processing performed by the control unit 31 of the image interpretation terminal 3 in response to the diagnosis support process is executed in cooperation with the control unit 31 and the image interpretation application stored in the storage unit 35.

In the diagnosis support process, when the control unit 21 receives operation information of a user operation on the viewer screen of the image interpretation terminal 3 through the communication unit 24, the control unit 21 determines whether a predetermined operation has been performed on the medical information of the target examination on the viewer screen. If a predetermined operation is performed, the operation information is stored in the RAM (step S1).
Here, the medical information includes medical images and interpretation reports. The predetermined operation is a predetermined operation used to specify the region of interest (the region interpreted by the doctor) in the medical image of the target examination. For example, as shown in FIG. Examples include operations such as panning and ROI settings, writing information to an interpretation report, and attaching an image to an interpretation report.

Next, the control unit 21 determines whether or not the user operation has been completed on the viewer screen based on the operation information from the image interpretation terminal 3 (step S2).
For example, the control unit 21 determines that the user operation has ended when a predetermined timing has arrived, such as when an end button or the like is pressed on the viewer screen.
If it is determined that the user operation has not been completed (step S2; NO), the control unit 21 returns to step S1.

If it is determined that the user operation has ended on the viewer screen (step S2; YES), the control unit 21 removes areas outside the region of interest from the medical image of the target examination based on the operation information of the predetermined operation stored in the RAM. The area is specified (step S3).
That is, based on a predetermined operation performed on the medical information of the target test, a region of interest in the medical image of the target test is identified, and a region outside the specified region of interest in the medical image is defined as a region outside the region of interest. Identify. AI may be used to specify areas outside this region of interest.

In step S3, for example, if the target examination includes a plurality of series and a predetermined operation is performed only on the medical images of some series, the control unit 21 first performs the above-mentioned predetermined operation. The area outside the region of interest is identified for the series of medical images in which the process was performed.
For medical images for which the region of interest could not be identified, the entire medical image is identified as an area outside the region of interest. This is based on the idea that "the region of interest is not specified = the entire image is outside the region of interest."
For example, if the medical image to be examined is a series of medical images generated by CT, for medical images in which the region of interest has been identified through a predetermined operation, areas outside the identified region of interest in the medical image can be For medical images that are identified as areas outside the region of interest and whose regions of interest cannot be identified because, for example, a predetermined operation was not performed, the entire medical image is identified as an area outside the region of interest.

The control unit 21 also controls medical images outside the region of interest, such as medical images that have been compared and interpreted with medical images on which a predetermined operation has been performed, or medical images of a series of medical images of a target examination on which a predetermined operation has not been performed. Areas outside the region of interest are also specified for other medical images for which the area has not been specified. Here, other medical images are medical images of the same patient as the medical images in which areas outside the region of interest have been identified (referred to as identified medical images), and are acquired with a different modality than the identified medical images. The medical image is at least one of a medical image, a medical image taken at a different time, or a medical image taken using a different imaging method.
For example, if a predetermined operation has been performed on another medical image, a region of interest in the other medical image is identified based on the predetermined operation, and a region other than the specified region of interest in the other medical image is identified. Identify the area outside the area of interest. If the predetermined operation is not performed on another medical image, an area corresponding to the area outside the area of interest of the identified medical image (an area where a similar part to that outside the area of interest of the identified medical image is drawn) may be specified as a region outside the region of interest. Alternatively, the entirety of another medical image may be specified as a region outside the region of interest.

Next, the control unit 21 uses AI to perform abnormality detection from a region outside the region of interest in the medical image (step S4).
The control unit 21 inputs the medical image in which the region outside the region of interest has been identified and the position information of the region outside the region of interest to the detection unit 26 (or inputs the image of the region outside the region of interest), Abnormality detection is performed using AI in areas outside the region of interest of the input medical image.
When a region outside the region of interest for another medical image is specified, the control section 21 inputs the position information of the other medical image and the region outside the region of interest to the detection section 26, and Anomaly detection is performed using AI in a region outside the region of interest in a medical image.
Note that abnormality detection using AI does not necessarily have to be performed on the entire region outside the region of interest, and may be performed on a part of the region outside the region of interest. For example, a subject area in a medical image may be recognized, and areas outside the subject area may be excluded from detection processing.
When the abnormality detection is completed, the detection section 26 outputs the detection result to the control section 21 .

When the abnormality detection is completed, the control unit 21 transmits the detection result to the image interpretation terminal 3 through the communication unit 24, and causes the image interpretation terminal 3 to output the detection result (step S5).
For example, if no abnormality is detected by the AI, as shown in Figure 6(a), a message indicating that the AI has completed the interpretation of the image outside the region of interest, such as "AI has completed the interpretation of the image outside the region of interest." A notification is sent to the image interpretation terminal 3 by the communication unit 24 and displayed on the display unit 33. Further, as shown in FIG. 6(b), an image indicating where the region outside the region of interest in which the abnormality was detected is is generated, transmitted to the image interpretation terminal 3 by the communication unit 24, and displayed on the display unit 33. . When an abnormality is detected by AI, as shown in Figure 7 (a), the AI will send a message outside the region of interest, such as "An abnormality was found as a result of AI interpretation outside the region of interest. Please check." The communication unit 24 sends a notification that an abnormality has been detected to the image interpretation terminal 3 and displays it on the display unit 33. Further, as shown in FIG. 7(b), the location where an abnormality is detected on the medical image is displayed surrounded by annotation A.

As explained above, according to the server device 2 of the diagnosis support system 100, the control unit 21 stores in the RAM a predetermined operation performed by the user on the medical information of the target test, and stores the stored target test Based on the user's predetermined operation on the medical information, an area outside the region of interest is specified from at least one medical image of the target examination, and the detection unit 26 performs abnormality detection on the specified area using artificial intelligence.
Therefore, abnormalities are automatically detected from outside the region of interest, which is not the subject of image interpretation by the doctor, so even if the abnormality exists in an area outside the region of interest, that is, an unintended area, it will assist in finding the abnormality. It is possible to efficiently prevent oversights in image interpretation. Further, since the detection unit 26 detects only the area outside the region of interest rather than the entire image, it is possible to efficiently prevent oversights with a small processing load.

Further, since the detection result by the detection unit 26 is output to the image interpretation terminal 3, a user such as a doctor can confirm whether or not an abnormality exists in a region outside the region of interest of the medical image.

Further, for example, the control unit 21 specifies a region of interest from at least one medical image of the target examination based on a user's predetermined operation on the medical information of the target examination stored in the RAM, and A region outside the region of interest is identified as a region outside the region of interest.
Therefore, areas outside the region of interest can be identified based on the operations that the user normally performs to identify the region of interest during image interpretation, so areas outside the region of interest can be identified without requiring the user to perform any special operations. It becomes possible to do so.

Further, with respect to other medical images of the same patient in which areas outside the region of interest have not been specified, the control unit 21 specifies the entire other medical image as outside the area of interest, and causes the detection unit 26 to identify the entire other medical image. Since abnormality detection is performed using artificial intelligence, for example, when performing comparative interpretation, the user can easily interpret other medical images, improving convenience.

Further, regarding other medical images of the same patient in which a region outside the region of interest has not been specified, the control unit 21 specifies, as outside the region of interest, an area corresponding to the region identified as outside the region of interest in the medical image, Since the detection unit 26 uses artificial intelligence to detect abnormalities in regions outside the region of interest in other medical images, for example, when performing comparative interpretation, the user can interpret the corresponding portions of other medical images. can be done easily, improving convenience.

For example, the control unit 21 controls the region of interest in at least one medical image of the same patient acquired with different modalities, medical images acquired at different times, or medical images photographed with different imaging methods. Since external areas are automatically identified and abnormalities detected, users can easily interpret these medical images, improving convenience.

Furthermore, if the area outside the area of interest cannot be identified from the medical image based on a predetermined operation stored in the RAM, the control unit 21 specifies the entire medical image as an area outside the area of interest. Even if the outside area cannot be specified, oversight can be prevented.

Note that the content described in the above embodiment is a preferred example of the present invention, and the present invention is not limited thereto.

For example, in the above embodiment, the present invention has been described as being applied to a diagnosis support system that includes the server device 2 and the image interpretation terminal 3, but the present invention may also be applied to a system that includes a console, an electronic medical record device, etc. . Further, the storage means, identification means, detection means, and output means of the present invention may be provided in any device constituting the system.
Moreover, the storage means, identification means, detection means, and output means of the present invention may be included in one diagnostic support device. For example, the control unit 31 of the image interpretation terminal 3 as a diagnostic support device, in cooperation with a program stored in the storage unit 35, controls a user's predetermined operation performed on the medical information of the target examination using the operation unit 32. and identifies a region outside the region of interest from at least one medical image of the target examination based on a predetermined operation performed by the user, and applies artificial intelligence to the region outside the region of interest identified in the medical image. The configuration may be such that abnormality detection is performed using the above-mentioned method, and the detection results are outputted on the display unit 33 or the like. Alternatively, a configuration may be adopted in which the received predetermined operation is temporarily stored in the RAM, and a region outside the region of interest in the medical image is specified based on the stored operation information. Further, the present invention is not limited to the image interpretation terminal 3, but may be applied to devices capable of displaying medical information, such as a console, a PACS server device, an electronic medical record device, etc.

Further, in the above description, an example is disclosed in which a hard disk, a semiconductor nonvolatile memory, etc. are used as a computer-readable medium for the program according to the present invention, but the present invention is not limited to this example. As other computer-readable media, it is possible to apply a portable recording medium such as a CD-ROM. Further, a carrier wave is also applied as a medium for providing data of the program according to the present invention via a communication line.

In addition, the detailed configuration and detailed operation of each device constituting the diagnosis support system can be changed as appropriate without departing from the spirit of the present invention.

100 Diagnosis support system 1 Image generation device 2 Server device 21 Control section 22 Operation section 23 Display section 24 Communication section 25 Storage section 251 Image DB
252 Additional information DB
26 Detection unit 27 Bus 3 Image interpretation terminal 31 Control unit 32 Operation unit 33 Display unit 34 Communication unit 35 Storage unit 36 Bus

Claims (9)

a storage means for storing a predetermined operation performed by a user on the medical information of the target test;
identification means for identifying an area outside the region of interest from at least one medical image of the target examination based on the user's predetermined operation on the medical information of the target examination stored in the storage means;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the specifying means;
A diagnostic support system equipped with The diagnosis support system according to claim 1, further comprising an output means for outputting a detection result by the detection means. The specifying means specifies a region of interest from at least one medical image of the target test based on the user's predetermined operation on the medical information of the target test stored in the storage unit, and The diagnostic support system according to claim 1 or 2, wherein a region outside the specified region of interest is specified as a region outside the region of interest. The specifying means further specifies, with respect to another medical image of the same patient as the medical image, in which a region outside the region of interest has not been specified, the entire other medical image to be outside the region of interest;
4. The diagnosis support system according to claim 1, wherein the detection means performs abnormality detection on the entire other medical image using artificial intelligence. The specifying means further determines, with respect to another medical image of the same patient as the medical image, in which a region outside the region of interest has not been specified, a region corresponding to the region identified as outside the region of interest in the medical image. Identified as outside the area,
4. The method according to claim 1, wherein the detecting means performs abnormality detection using artificial intelligence only in a region of the other medical image that is outside the region of interest of the other medical image. Diagnostic support system. 4. The other medical image is at least one medical image among medical images acquired with a different modality, medical images acquired at different times, or medical images photographed with a different imaging method. Or the diagnostic support system described in 5. The identifying means identifies the entire medical image as a region outside the region of interest when the region outside the region of interest cannot be identified from the medical image based on the predetermined operation stored in the storage means. The diagnostic support system according to any one of Items 1 to 6. identification means for accepting a predetermined operation performed by a user on the medical information of the target examination, and specifying an area outside the region of interest from at least one medical image of the target examination based on the predetermined operation;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the specifying means;
A diagnostic support device comprising: computer,
storage means for storing predetermined operations performed by the user on the medical information of the target test;
identification means for identifying an area outside the region of interest from at least one medical image of the target examination based on the user's predetermined operation on the medical information of the target examination stored in the storage means;
Detection means for performing abnormality detection using artificial intelligence only in a region of the medical image that is outside the region of interest of the medical image specified by the specifying means;
A program to function as

Images