Endoscopic image diagnosis assistance processor, endoscopic image diagnosis assistance processor operation method, and endoscopic image diagnosis assistance processor program

Abstract

[Problem] To provide an endoscopic image diagnosis assistance processor 2 for stably identifying a to-be-observed organ which is being observed by means of an endoscope 9. [Solution] This endoscopic image diagnosis assistance processor 2 comprises: a site inference unit 13 that uses a first model for an endoscopic image to infer a certainty factor of each of a plurality of site images; a site identification unit 14 that identifies a to-be-observed site of the endoscopic image, on the basis of a predetermined condition using the certainty factor; an organ identification unit 14 that identifies a first to-be-observed organ corresponding to the to-be-observed site; a lesion detection unit 20 that uses a plurality of second models, which are trained on a plurality of lesion images, as teacher data, for each of a plurality of organs, to infer a lesion in the aforementioned image; an organ specification unit 17 that enables a user to specify a second model of a second to-be observed organ among the plurality of second models; and a warning unit 18 that issues a warning on the basis of a predetermined condition when the first to-be-observed organ and the second to-be-observed organ are different.

Description

Endoscopic image diagnosis support processor, operating method of endoscopic image diagnosis support processor, and program for endoscopic image diagnosis support processor , ,

[0004] , ,

[0003] ,

[0006] ,

[0005] ,

[0001] Embodiments of the present invention relate to an endoscopic image diagnosis support processor that supports a user's judgment of an organ being observed, an operating method of the endoscopic image diagnosis support processor that supports a user's judgment of an organ being observed, and a program for the endoscopic image diagnosis support processor that supports a user's judgment of an organ being observed.

[0002] For an image captured using an endoscope, computer-aided detection (CADe: Computer-Aided Detection) indicating the position of a lesion candidate and computer-aided diagnosis (CADx: Computer-Aided Diagnosis) indicating discrimination information of the lesion candidate are known (hereinafter, CADe and CADx are collectively referred to as "CAD"). In CAD, for example, a machine learning model obtained by performing deep learning or the like using a plurality of images as teacher data is used.

[0003] International Publication No. WO2022 / 181748 discloses a technique for generating a plurality of models each adapted to a plurality of organs by learning using images of each of the plurality of organs. For example, a plurality of types of models such as a pharyngeal model learned using only pharyngeal images and an esophageal model learned using only esophageal images are created. Then, by using a model adapted to the organ being observed with the endoscope from among the plurality of types of models, the accuracy as CAD can be increased.

[0004] When the operator designates the organ being observed by himself / herself, if the designated organ is incorrect, there is a risk that an accurate diagnosis cannot be made.

[0005] International Publication No. WO2022 / 181748 Japanese Unexamined Patent Application Publication No. 2007-151809

[0006] Embodiments of the present invention aim to provide an endoscopic image diagnostic support processor that reliably identifies an organ being observed with an endoscope, a method for operating the endoscopic image diagnostic support processor that reliably identifies an organ being observed with an endoscope, and a program for the endoscopic image diagnostic support processor that allows a computer to reliably identify an organ being observed with an endoscope.

[0007] The endoscopic image diagnostic support processor according to an embodiment of the present invention comprises: a site inference unit that, for a plurality of endoscopic images input in succession, uses a first model trained with multiple images of multiple sites as training data to infer the confidence level of each endoscopic image being an image of the respective site; a site identification unit that identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level; an organ identification unit that identifies a first observation organ corresponding to the observation site; a lesion detection unit that infers lesions in the image using a plurality of second models trained with multiple lesion images for each of the multiple organs as training data; an organ designation unit that allows the user to specify a second observation organ; and a warning unit that issues a warning based on predetermined conditions when the first observation organ and the second observation organ are different.

[0008] The operation method of the endoscopic image diagnostic support processor according to an embodiment of the present invention involves, for a plurality of endoscopic images that are input in succession, using a first model that has been trained using multiple images of multiple body parts as training data, inferring the confidence level for each endoscopic image to be an image of the respective body part, identifying the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifying the first observation organ corresponding to the observation site, and issuing a warning based on predetermined conditions if the second model for a specified second observation organ differs from the second model corresponding to the first observation organ among a plurality of second models that have been trained using multiple lesion images for each of the multiple organs as training data.

[0009] The program for the endoscopic image diagnostic support processor according to an embodiment of the present invention, in response to a plurality of endoscopic images that are input sequentially, uses a first model that has learned multiple images of multiple body parts as training data to infer the confidence level of each endoscopic image that is an image of one of the multiple body parts, identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifies the first observation organ corresponding to the observation site, and causes the computer to execute a process to issue a warning based on predetermined conditions if the second model for a specified second observation organ and the second model corresponding to the first observation organ are different among a plurality of second models that have learned multiple lesion images for each of the multiple organs as training data.

[0010] According to embodiments of the present invention, an endoscopic image diagnostic support processor that reliably identifies an organ being observed with an endoscope, a method for operating the endoscopic image diagnostic support processor that reliably identifies an organ being observed with an endoscope, and a program for the endoscopic image diagnostic support processor that allows a computer to reliably identify an organ being observed with an endoscope are provided.

[0011] Figure 1 is a configuration diagram of the endoscopic image diagnostic support processor of the first embodiment. Figure 2 is a flowchart of the operation method of the endoscopic image diagnostic support processor of the first embodiment. Figure 3 is an example of the screen of the display device of the embodiment. Figure 4 is a flowchart of the operation method of the endoscopic image diagnostic support processor of modification 1 of the first embodiment. Figure 5 is a flowchart of the operation method of the endoscopic image diagnostic support processor of modification 2 of the first embodiment. Figure 6 is a configuration diagram of the endoscopic image diagnostic support processor of the second embodiment. Figure 7 is a flowchart of the operation method of the endoscopic image diagnostic support processor of the second embodiment. Figure 8 is a flowchart of the operation method of the endoscopic image diagnostic support processor of modification 2 of the second embodiment.

[0012] Embodiments of the present invention will be described below with reference to the drawings. The drawings based on the embodiments are schematic. The illustration and reference numerals of some components have been omitted.

[0013] <First Embodiment> As shown in Figure 1, the endoscopic image diagnostic support processor 2 (hereinafter referred to as "processor 2") of this embodiment, together with the endoscope 9, display device 30, notification unit 31, and storage device 32, constitutes the endoscopic system 1.

[0014] The camera unit 8 of the endoscope 9 captures images of the inside of the subject's body. The endoscope 9 outputs image data at a predetermined frame rate. The user can capture still images, for example, by operating the operation buttons 7 on the control panel (not shown) of the endoscope 9. The endoscope 9 may be either a flexible or rigid endoscope.

[0015] The processor 2 has an observed organ determination device 10 (hereinafter referred to as "determination device 10").

[0016] The determination device 10 includes a CPU 11, an image input unit 12, a site inference unit 13, a site identification unit 14, an organ identification unit 15, a memory 16, an organ designation unit 17, a warning unit 18, and a warning stop instruction unit 19. The processor 2 is also connected to the server 40.

[0017] The CPU 11 is a system controller that controls the entire processor 2. The image input unit 12 processes images taken by the endoscope 9 at a predetermined frame rate. Alternatively, the image input unit 12 may receive endoscope images processed by an image processing processor (not shown) or a server 40 from the image data from the endoscope 9.

[0018] The site inference unit 13 is a first AI processing unit that uses a first model, which has been trained using multiple images of multiple sites as training data, to infer the confidence level of each of the multiple sites in the endoscopic image input from the endoscope. The site identification unit 14 identifies the observation site, which is a site in the endoscopic image, from among the multiple sites based on predetermined conditions using the confidence level P.

[0019] The organ identification unit 15 identifies the first organ to be observed that corresponds to the observation site. The memory 16 stores various data, such as the correspondence between the observation site and the first organ to be observed.

[0020] The organ selection unit 17 is a user interface that allows the user to specify a second organ to be observed, which the user has determined to be the organ being observed. The warning unit 18 issues a warning if the first organ to be observed, which has been identified by the organ identification unit 15, is different from the second organ to be observed, which the user has specified. The warning stop instruction unit 19 issues a stop signal to stop the warning based on the user's operation.

[0021] Furthermore, at least one of the multiple configurations of the processor 2 may consist of internal circuits of semiconductor elements processed by software, or dedicated hardware circuits, or it may include both internal circuits of semiconductor elements and dedicated hardware circuits. At least one of the functional parts of the processor 2 may also be a component of the server 40. For example, the lesion detection unit 20 may be an operating unit of the server 40.

[0022] For example, the body part inference unit 13 may include an FPGA (Field Programmable Gate Array) with hardware circuits and a GPU (Graphics Processing Unit) with software circuits. Also, the body part identification unit 14, organ identification unit 15, etc., may be part of a CPU 11 that processes data using software.

[0023] The display device 30, which displays the endoscopic image, is a monitor such as a liquid crystal display. The notification unit 31 informs the user of information from the processor 2 using sound, light, etc. The notification unit 31 may be part of the display device 30.

[0024] The storage device 32 is a non-temporary storage device (e.g., a magnetic disk, optical disk, or HDD) that stores the program of the processor 2. The storage device 32 may also be configured as a server 40 connected to the processor 2 via an internet connection or the like.

[0025] In processor 2, the determination device 10 can efficiently and stably identify the organ being observed in the endoscopic image.

[0026] <Processor Operation Method> The operation method of processor 2 will be explained according to the flowchart in Figure 2.

[0027] <Step S10> Image Input The endoscopic system 1 of this embodiment performs an examination of the upper gastrointestinal tract. The insertion part of the endoscope 9 is inserted through the oral cavity or nasal cavity of the subject and enters the pharynx. For example, multiple endoscopic images from the endoscope 9 or the image processing processor are input to the image input unit 12 of the processor 2 in succession at a predetermined frame rate. Processing is not performed if no images are input.

[0028] The image input unit 12 performs image processing on images directly input from the endoscope 9 (for example, RAW images) for display on the display device 30. If an image processed by an image processing processor is input, no image processing is required.

[0029] <Step S20> The endoscopic image is displayed on the endoscopic image display device 30.

[0030] <Step S30> Site Inference The site inference unit 13 is a first AI processing unit that uses a first model to infer the confidence level PN (N=1 to n) that the input endoscopic image is an image of one of n different sites. The value of the confidence level P is in the range of (0 to 1). (Confidence level P=0) means a probability of 0%, and (Confidence level P=1) means a probability of 100%.

[0031] Furthermore, the region inference unit 13 does not need to infer all images that are continuously input at a predetermined frame rate. For example, the region inference unit 13 processes only 20 fps (20 frames / second) image data from 120 fps (120 frames / second) image data.

[0032] The machine learning model in the AI ​​processing unit is a deep neural network (DNN) that performs deep learning with multiple hidden layers. The model can also be a convolutional neural network (CNN), R-CNN (Regions with CNN features) which utilizes CNN, or FCN (Fully Convolutional Networks), etc.

[0033] The first model is trained using images of n different parts of the upper digestive tract as training data. As shown in Table 1 below, in this embodiment, n = 15.

[0034]

[0035] Each of the 15 body parts corresponds to one of the m types of organs. As shown in Table 2 below, in this embodiment, m = 6.

[0036]

[0037] The number of types of body parts n is, for example, between 2 and 100. The number of types of organs m is (m < n), for example, between 2 and 20.

[0038] The site inference unit 13 estimates the probability (confidence level P1 to P15) that the endoscopic image represents one of 15 different sites. For example, the site inference unit 13 outputs an inference result consisting of 15 confidence levels (P1 to P15), such as a confidence level P12 of 0.1 for site 12 (antrum) and a confidence level P14 of 0.6 for site 14 (duodenal bulb).

[0039] <Step S40> The part identification unit 14 identifies the observation area, which is a part of the endoscopic image, from among multiple parts based on predetermined conditions using confidence level P.

[0040] In the processor 2 of this embodiment, among the 15 types of parts (parts 1 to 15), the part with the highest confidence level P is designated as the observed part X (confidence level PX).

[0041] <Step S50> The organ identification unit 15 identifies the first observed organ Y that corresponds to the observed site X, for example, using a correspondence table (Table 2) stored in the memory 16.

[0042] The identified observation site X and the first observation organ Y may be displayed on the display device 30.

[0043] <Step S60> When the user who has viewed the endoscopic image displayed on the display device 30 in Step S20 with the second observation organ specified determines that the organ in the image being observed has changed to the second observation organ Z, a signal for designating the second observation organ Z is generated from the organ designation unit 17 based on the user's operation (S60: YES).

[0044] For example, in FIG. 3, when the icon of the second observation organ Z is clicked among the plurality of organs graphically displayed on the display device 30 using the figure 30D, the icon is highlighted and a second observation organ designation signal is generated.

[0045] While the user determines that the organ in the image being observed has not changed, a signal for designating a new second observation organ Z is not generated (S60: NO).

[0046] <Step S70> Second Observation Organ Setting Based on the signal for designating the second observation organ Z, the second observation organ Z is set.

[0047] Note that the processing of Steps S30 to S50 and the processing of Steps S60 to S70 may be performed in parallel, or the processing of Steps S60 to S70 may be performed prior to the processing of Steps S30 to S50.

[0048] <Step S80> Match? Whether the first observation organ Y specified by the organ identification unit 15 and the second observation organ Z specified by the organ designation unit 17 are the same is determined, for example, by the CPU 11.

[0049] When the first observation organ Y and the second observation organ Z match (S80: YES), the processing from Step S90 is performed. When the first observation organ Y and the second observation organ Z do not match (S80: NO), the processing from Step S110 is performed.

[0050] <Step S90> Warning in Progress? Whether the warning unit 18 is already generating a warning is determined, for example, by the CPU 11. When the warning is not in progress, the processing from Step S10 is performed. On the other hand, when the warning is in progress, in Step S100, the warning is stopped and the processing from Step S10 is performed.

[0051] <Step S110> Is a warning issued? If no warning is issued, the process from step S120 is performed. On the other hand, if a warning is issued, the process from step S130 is performed.

[0052] <Step S120> Timer TM1 Activation Timer TM1 starts measuring time. The measurement time of Timer TM1 is reset when activated (TM = 0).

[0053] <Step S130> TM1 ≥ T1 It is determined whether the time measured by timer TM1 is greater than or equal to the first determination value T1. If the time exceeding the first determination value T1 has not elapsed (NO), the process from step S10 is performed. If the time exceeding the first determination value T1 has elapsed (YES), the process in step S140 is performed.

[0054] The first judgment value T1 can be set as appropriate, but for example, it is 0.5 seconds.

[0055] <Step S140> Warning Generation The warning unit 18 generates a warning. The warning is displayed on the display device 30, for example. In the example shown in Figure 3, the icon for the second observed organ Z, which is graphically displayed using a figure 30D, is "stomach," while the first observed organ Y, which is displayed as text 30C, is "esophagus." Therefore, the stomach icon (30D) and the text for esophagus (30C) are displayed in red or flash.

[0056] Warnings are not limited to graphics or other images displayed on the display device 30; they may also include sounds, vibrations, or light.

[0057] The user can recognize that the first observed organ Y, which is automatically identified by the warning unit 18, is different from the second observed organ Z, which they themselves have determined to be the correct organ, thus enabling them to reliably identify the observed organ.

[0058] As described above, the endoscopic image diagnostic support processor of this embodiment comprises: a site inference unit that, for multiple endoscopic images input in succession, uses a first model trained on multiple images of multiple sites as training data to infer the confidence level of each endoscopic image being an image of one of the multiple sites; a site identification unit that identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level; an organ identification unit that identifies a first observation organ corresponding to the observation site; an organ designation unit that designates a second observation organ; and a warning unit that issues a warning based on predetermined conditions when the first observation organ and the second observation organ are different.

[0059] The operation method of the endoscopic image diagnostic support processor of this embodiment involves, for multiple endoscopic images that are input in succession, using a first model that has learned multiple images of multiple body parts as training data, inferring the confidence level of each endoscopic image to be an image of one of the multiple body parts, identifying the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifying a first observation organ corresponding to the observation site, and if a second observation organ is specified and the first observation organ and the second observation organ are different, issuing a warning based on predetermined conditions.

[0060] The program for the endoscopic image diagnostic support processor of this embodiment uses a first model, which has been trained using multiple images of multiple body parts as training data, to infer the confidence level of each endoscopic image being an image of one of the multiple body parts, based on predetermined conditions using the confidence level, to identify the observation site of the endoscopic image, to identify the first observation organ corresponding to the observation site, and to cause the computer to execute a process to issue a warning based on predetermined conditions if a second observation organ is specified and the first observation organ and the second observation organ are different.

[0061] The non-temporary storage medium of this embodiment stores a program for an endoscopic image diagnostic support processor that, for multiple endoscopic images input sequentially, uses a first model trained with multiple images of multiple body parts as training data to infer the confidence level of each endoscopic image being an image of one of the multiple body parts, identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifies a first observation organ corresponding to the observation site, and if a second observation organ is specified and the first observation organ and the second observation organ are different, causes the computer to execute a process to issue a warning based on predetermined conditions.

[0062] The non-temporary storage medium, the storage device 32, may store only a part of the program. Furthermore, the program may be distributed or provided via a communication network. Users can execute all or part of the processing, and thus perform the processing of the processor described above, by installing the program from the storage medium to their computer, or by downloading the program via the communication network and installing it to their computer.

[0063] <Modifications of the First Embodiment> The processor 2A-2D and determination device 10A-10D of the modified embodiments and embodiments described below are similar to the processor 2 and determination device 10 and have the same effects. For this reason, in the following, components with the same function as the processor 2 and determination device 10 are denoted by the same reference numerals, and are not shown or described. In the flowchart, the same processes are denoted by the same step numbers, and are not described.

[0064] <Modification 1 of the First Embodiment> As shown in the flowchart of Figure 4, in the processor 2A of this modification, the warning unit 18 generates a warning based on a counter, not on the timer TM1.

[0065] Specifically, in step S122, the count K is initialized (K = 0), and in step S124, the number of images in which the first observed organ and the second observed organ do not match is counted (K = K + 1). If the count K is greater than or equal to the second determination value T2, the warning unit 18 issues a warning.

[0066] The second judgment value T2 can be set as appropriate, but for example, it is 30.

[0067] <Modification 2 of the First Embodiment> As shown in the flowchart of Figure 5, in the processor 2B of this modification, the warning unit 18 stops the generated warning even if the first observed organ and the second observed organ do not match.

[0068] For example, if a warning is generated in step 140, in step S150, timer TM2 is started, and it is determined whether the time measured by timer TM2 is equal to or greater than the third determination value T3 (step S160). If a time exceeding the third determination value T3 has elapsed since the warning was generated (S160: YES), the warning is stopped in step S100.

[0069] Furthermore, even if a time exceeding the third determination value T3 has not elapsed since the warning was issued (S160: NO), if a stop signal is received from the warning stop instruction unit 19 (S170: YES), the warning unit 18 stops the warning in step S100.

[0070] Furthermore, if the warning is stopped in step S100 via steps S160 and S170, it is preferable that the warning will not be issued again, even if the first observed organ Y and the second observed organ Z are different, until a new first observed organ Y is identified in step S50 or a new second observed organ Z is specified by the user in step S60.

[0071] While warnings are beneficial to users, if they continue after the user has recognized a difference in the observed organ, the user may find them annoying, potentially leading to a decrease in work efficiency.

[0072] Processor 2B offers better performance than Processor 2.

[0073] <Second Embodiment> As shown in Figure 6, the processor 2C of this embodiment includes a lesion detection unit 20 for detecting lesions in the organ being observed.

[0074] The lesion detection unit 20 is a second AI processing unit that infers lesions in images using multiple second models that have been trained using multiple lesion images for multiple organs as training data. In other words, the lesion detection unit 20 has a second model corresponding to each of the multiple organs.

[0075] The lesion detection unit 20 includes a nasal cavity CAD (second model 1) 21, an oral cavity CAD (second model 2) 22, a pharyngeal CAD (second model 3) 23, an esophageal CAD (second model 4) 24, a gastric CAD (second model 5) 25, a duodenal CAD (second model 6) 26, and an ON / OFF control unit 29 for controlling the activation / deactivation of each model. Each model may also have its own ON / OFF control unit.

[0076] <Step S106> Lesion detection processing: The activated CAD of the lesion detection unit 20 infers the lesion from the endoscopic image. For example, the activated gastric CAD 30 performs inference using the second model 5. The inference result is output as a confidence level PP (PP = 0 to 1) inferred by the CAD corresponding to the observed organ Y.

[0077] For example, the confidence level PP1 for normal tissue is output as 0.66, and the confidence level PP2 for diseased tissue is output as 0.34. The lesion detection unit 20 may output either confidence level PP1 or confidence level PP2.

[0078] Furthermore, when processor 2 starts up, i.e., at the start of the examination, there are no organs identified as being observed, and therefore no CAD is running. For this reason, lesion detection processing is not performed.

[0079] <Step S108> If the confidence level PP2 of the displayed lesion tissue is above a predetermined value, a warning message is displayed on the display device 30 showing the first and second observed organs, for example, on the endoscopic image 30A.

[0080] For example, the endoscopic image 35A may have candidate lesions superimposed on it, or the confidence PP2 may be displayed numerically or in color. For example, candidate lesions with a confidence PP2 of 0.7 or higher may be displayed in red and superimposed on the endoscopic image 35A.

[0081] In addition, if the system is not in AUTO mode (off: manual operation), a second model corresponding to the second observation organ Z specified by the user may be activated, and lesion detection processing may be performed. In this case, the user must appropriately change the observation organ specification during the examination. If the user forgets to switch the observation organ, lesion detection processing will be performed by a second model that does not correspond to the organ being observed.

[0082] Processor 2C can perform appropriate lesion detection processing even if the user forgets to switch the observed organ, as it will generate a warning.

[0083] Furthermore, the "organ CAD" may include a "boundary CAD" that performs lesion detection processing on images containing multiple organs. For example, if the confidence level of site 4 (esophagus) and the confidence level of site 5 (esophagogastric junction) are both above a predetermined confidence level P, the boundary organ including the esophagus and stomach may be identified, and the corresponding boundary CAD may infer lesions in the endoscopic image. Alternatively, one organ may be divided into multiple organs for processing purposes, and any of the multiple CADs corresponding to one organ may infer lesions in the endoscopic image.

[0084] <Modification of the Second Embodiment> In the processor 2D of this modification, the organ identification unit 15 is a second model identification unit that identifies a second model corresponding to each of a plurality of organs. The organ designation unit 17 is a second model designation unit in which the user designates a second model corresponding to each of a plurality of organs.

[0085] In other words, in the flowchart of the processor 2D shown in Figure 8, in step S50, the first organ to be observed is identified, and the second model corresponding to the identified first organ to be observed is identified. In step S65, when the second model is specified based on user operation, a signal to specify the second model is generated from the organ specification unit 17, which is the second model specification unit (S65: YES). In step S70, the specified second model is set as the model for lesion detection.

[0086] Then, in step S80, it is determined whether the second model identified by the inference of the first model matches the second model specified by the user.

[0087] The warning unit 18 of the determination device 10D issues a warning based on predetermined conditions if the activated second model is different from the designated second model.

[0088] The operation method of processor 2D in this modified example involves, for multiple endoscopic images that are input sequentially, using a first model that has been trained using multiple images of multiple body parts as training data, inferring the confidence level for each endoscopic image to be an image of one of the multiple body parts, identifying the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifying the first observation organ corresponding to the observation site, and issuing a warning based on predetermined conditions if the second model for a specified second observation organ differs from the second model corresponding to the first observation organ among multiple second models that have been trained using multiple lesion images for each of the multiple organs as training data.

[0089] The range of values ​​described above is not limited to that range and can be increased or decreased as appropriate. Furthermore, the present invention is not limited to the embodiments and modifications described above, and various changes and modifications can be made as long as they do not alter the essence of the present invention.

[0090] 1... Endoscope system 2, 2A-2D... Endoscope image diagnostic support processor (processor) 7... Operation buttons 8... Camera unit 9... Endoscope 10, 10A-10D... Observed organ determination device (determination device) 11... CPU 12... Image input unit 13... Site inference unit 14... Site identification unit 15... Organ identification unit 16... Memory 17... Organ designation unit 18... Warning unit 19... Organ identification unit 20... Warning stop instruction unit 21-26... Second model 29... ON / OFF control unit 30... Display device 31... Notification unit 32... Storage device

Claims

1. An endoscopic image diagnostic support processor comprising: a site inference unit that, for multiple endoscopic images input in succession, uses a first model trained on multiple images of multiple sites as training data to infer the confidence level of each endoscopic image being an image of one of the multiple sites; a site identification unit that identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level; an organ identification unit that identifies a first observation organ corresponding to the observation site; a lesion detection unit that infers lesions in the image using multiple second models trained on multiple lesion images for each of the multiple organs as training data; an organ designation unit that allows the user to specify a second observation organ; and a warning unit that issues a warning based on predetermined conditions when the first observation organ and the second observation organ are different.

2. The endoscopic image diagnostic support processor according to claim 1, wherein the warning unit issues a warning if the first observed organ and the second observed organ are different for a period of time equal to or greater than the first determination value.

3. The endoscopic image diagnostic support processor according to claim 1, wherein the warning unit issues a warning when the first observed organ and the second observed organ are different in a plurality of endoscopic images that are equal to or greater than a second determination value.

4. The endoscopic image diagnostic support processor according to claim 1, wherein the warning unit stops the warning if a period of time equal to or greater than a third determination value has elapsed since the warning was issued.

5. The endoscopic image diagnostic support processor according to claim 4, wherein the warning unit does not issue a warning even if the first observed organ and the second observed organ are different, until the first observed organ is changed and / or a new second observed organ is specified.

6. The endoscopic image diagnostic support processor according to claim 1, comprising a warning stop instruction unit into which an instruction to stop the warning is input, wherein the warning unit stops the warning based on the instruction of the warning stop instruction unit.

7. The endoscopic image diagnostic support processor according to claim 6, wherein the warning unit does not issue a warning even if the first observed organ and the second observed organ are different, until the first observed organ is changed and / or a new second observed organ is specified.

8. The endoscopic image diagnostic support processor according to claim 1, further comprising a lesion detection unit that infers lesions in an image using a plurality of second models that have been trained using a plurality of lesion images for a plurality of organs as training data, wherein the lesion detection unit uses a second model from the plurality of second models that corresponds to the second observed organ.

9. The endoscope image diagnostic support processor according to claim 1, wherein the site identification unit identifies the site with the highest confidence level as the observation site.

10. The endoscopic image diagnostic support processor according to claim 1, further comprising a notification unit that notifies the user of at least one of the first or second organs to be observed.

11. Endoscopic image diagnostic support processor comprising: a site inference unit that, for multiple endoscopic images input in succession, uses a first model trained on multiple images of multiple sites as training data to infer the confidence level of each endoscopic image being an image of one of the multiple sites; a site identification unit that identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level; an organ identification unit that identifies a first observation organ corresponding to the observation site; a lesion detection unit that infers lesions in the image using multiple second models trained on multiple lesion images for each of the multiple organs as training data; an organ designation unit that allows the user to specify one of the multiple second models; and a warning unit that issues a warning based on predetermined conditions when the second model corresponding to the first observation organ among the multiple second models is different from the one second model designated by the organ designation unit.

12. A method for operating an endoscopic image diagnostic support processor that, in response to multiple endoscopic images input in succession, uses a first model trained with multiple images of multiple body parts as training data to infer the confidence level for each endoscopic image to be an image of one of the multiple body parts; identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level; identifies a first observation organ corresponding to the observation site; and, if the second model for a specified second observation organ differs from the second model corresponding to the first observation organ among multiple second models trained with multiple lesion images for each organ as training data, the processor issues a warning based on predetermined conditions.

13. A program for an endoscopic image diagnostic support processor that, in response to multiple endoscopic images input in succession, uses a first model trained on multiple images of multiple body parts as training data to infer the confidence level for each endoscopic image to be an image of one of the multiple body parts, identifies the observation site of the endoscopic image based on predetermined conditions using the confidence level, identifies a first observation organ corresponding to the observation site, and, if the second model for a specified second observation organ differs from the second model corresponding to the first observation organ among multiple second models trained on multiple lesion images for each organ as training data, the program causes the computer to execute a process to issue a warning based on predetermined conditions.

Images