Endoscope image processing device, endoscope image processing method, endoscope image processing program, non-volatile storage medium, and endoscope system

Abstract

This endoscope image processing device includes: a recognition reception unit for receiving a recognition result of recognizing a lesion from an image; a basis information generation unit for generating basis information for the recognition result; a difficulty determination unit for determining recognition difficulty from a recognition difficulty determination material; and an output unit for outputting, to a display unit, the basis information if the difficulty determination unit determines that recognition is difficult.

Description

Endoscopic image processing device, endoscope image processing method, endoscope image processing program, non-volatile storage medium, and endoscope system

[0001] The present invention relates to an endoscopic image processing device, an endoscopic image processing method, an endoscopic image processing program, a nonvolatile storage medium, and an endoscopic system.

[0002] 2. Description of the Related Art Conventionally, as disclosed in Japanese Patent No. 7161979, there is known XAI (Explainable AI) that presents information on the basis of decisions made by AI (Artificial Intelligence).

[0003] For example, XAI may be applicable to medical image diagnosis support systems (CAD: Computer Aided Diagnosis / Detection) used in medical settings, such as for endoscopic diagnosis.

[0004] However, when XAI is applied to a medical image diagnosis support system, a large amount of decision-making basis information is frequently displayed on the screen, which may become cumbersome for endoscopists who are required to make accurate diagnoses in a short amount of time.

[0005] The present invention has been made in consideration of the above-mentioned circumstances, and aims to provide an endoscopic image processing device, an endoscopic image processing method, an endoscopic image processing program, a non-volatile storage medium, and an endoscopic system that can appropriately display judgment basis information.

[0006] An endoscopic image processing device according to one aspect of the present invention includes a recognition receiving unit that receives a recognition result of recognizing a lesion from an image, a basis information generating unit that generates basis information for the recognition result, a difficulty determining unit that determines the difficulty of recognition based on recognition difficulty determination material, and an output unit that outputs the basis information to a display unit when the difficulty determining unit determines that recognition is difficult.

[0007] Another aspect of the present invention is an endoscopic image processing device that includes a processor that receives a recognition result of a lesion recognized from an image, generates basis information for the recognition result, determines the difficulty of recognition based on at least one of the recognition result, the image, the basis information, or the imaged portion of the image, and outputs the basis information to a monitor if it determines that recognition is difficult.

[0008] In addition, an endoscopic image processing method of one aspect of the present invention receives a recognition result of recognizing a lesion from an image, generates basis information for the recognition result, determines the difficulty of recognition, and outputs the basis information to a monitor if it is determined that recognition is difficult.

[0009] In addition, an endoscopic image processing program according to one embodiment of the present invention causes a computer to execute the following processes: receiving a recognition result of a lesion recognized from an image; generating basis information for the recognition result; determining the difficulty of recognition; and outputting the basis information to a monitor if it is determined that recognition is difficult.

[0010] Furthermore, a non-volatile storage medium according to one aspect of the present invention is a non-volatile storage medium storing an endoscopic image processing program, which causes a computer to execute the following processes: receiving a recognition result of recognizing a lesion from an image; generating basis information for the recognition result; determining the difficulty of recognition; and outputting the basis information to a monitor if it is determined that recognition is difficult.

[0011] Also, an endoscopic system according to one aspect of the present invention includes an endoscope, a recognition result generating device connected to the endoscope, and the above-mentioned endoscopic image processing device connected to the endoscope and the recognition result generating device.

[0012] FIG. 1 is a block diagram showing the configuration of an endoscopic image processing device according to a first embodiment. FIG. 2 is a diagram showing an example of a display screen displayed on the display unit when recognition is easy. FIG. 3 is a diagram showing an example of a display screen displayed on the display unit when recognition is difficult. FIG. 4 is a flowchart showing an example of the operation of the endoscopic image processing device according to the first embodiment. FIG. 5 is a block diagram showing the configuration of an endoscopic image processing device according to a first modified example of the first embodiment. FIG. 6 is a block diagram showing the configuration of an endoscopic image processing device according to a second embodiment. FIG. 7 is a flowchart showing an example of the operation of the endoscopic image processing device according to the second embodiment. FIG. 8 is a block diagram showing the configuration of an endoscopic image processing device according to a first modified example of the second embodiment. FIG. 9 is a block diagram showing the configuration of an endoscopic image processing device according to a second modified example of the second embodiment. FIG. 10 is a block diagram showing the configuration of an endoscopic image processing device according to a second modified example of the second embodiment. FIG. 11 is a block diagram showing the configuration of an endoscopic image processing device according to a second modified example of the second embodiment.

[0013] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. (First Embodiment) Fig. 1 is a block diagram showing the configuration of an endoscopic image processing device according to a first embodiment. As shown in Fig. 1, the endoscopic image processing device 1 includes an endoscopic image input unit 11, a recognition result receiving unit 12, a difficulty determining unit 13, an output unit 14, a determination basis generating unit 15, and a display unit 16. Note that the display unit 16 may be a display device provided separately from the endoscopic image processing device 1.

[0014] An endoscopic image captured by an endoscope is input to the endoscopic image input unit 11. The endoscopic image input unit 11 outputs the input endoscopic image to the determination basis generating unit 15 and the display unit 16.

[0015] The endoscopic image is also input to a recognition result generating unit 17 provided outside the endoscopic image processing device 1. The recognition result generating unit 17 generates a recognition result including a lesion determination result (estimation result) and a confidence level from the endoscopic image, and outputs the recognition result to the recognition result receiving unit 12. The lesion determination result is, for example, a JNET classification, a NICE classification, or the like. The confidence level is information of high or low. The confidence level may also be a continuous value score.

[0016] The recognition result receiving unit 12 outputs the recognition result input from the recognition result generating unit 17 to the difficulty determining unit 13 and the display unit 16 .

[0017] The difficulty determination unit 13 determines the difficulty of recognizing the lesion from the recognition result. Specifically, the difficulty determination unit 13 determines that recognition is easy when the confidence level is high, and determines that recognition is difficult when the confidence level is low. Note that when the confidence level is a continuous value score, the difficulty determination unit 13 compares the confidence level with a predetermined threshold, and determines that recognition is easy when the confidence level is higher than the predetermined threshold, and determines that recognition is difficult when the confidence level is equal to or lower than the predetermined threshold.

[0018] In this way, the confidence level input from the recognition result receiving unit 12 becomes a recognition difficulty determination material for determining the recognition difficulty in the difficulty determining unit 13. The difficulty determining unit 13 outputs the determination result to the output unit 14.

[0019] The difficulty determination unit 13 may determine the recognition difficulty based on the determination result (estimation result) input from the recognition result receiving unit 12. For example, the difficulty determination unit 13 determines that recognition is difficult when the determination result is a predetermined disease name. The predetermined disease name is, for example, cancer, inflammatory bowel disease (IBD), gastritis, Helicobacter pylori infection, or the like.

[0020] Furthermore, the difficulty determining unit 13 may determine that recognition is difficult when the location where the lesion occurs is a location where it is difficult to find the lesion, such as squamous epithelium or the inside of a polyp.

[0021] Furthermore, the difficulty determination unit 13 may determine that recognition is difficult when the observation conditions make it difficult to find a lesion, such as when the size of the lesion is less than a predetermined value, when the protrusion of the lesion is less than a predetermined value, when the brightness of the endoscopic image is less than a predetermined value, or when blood, residue, or bubbles are reflected in the endoscopic image.

[0022] The output unit 14 determines whether or not to display the judgment basis information based on the judgment result of the difficulty judgment unit 13. If the difficulty judgment unit 13 determines that recognition is easy, the output unit 14 determines not to display the judgment basis information. On the other hand, if the difficulty judgment unit 13 determines that recognition is difficult, the output unit 14 determines to display the judgment basis information. The output unit 14 outputs the judgment result thus determined to the judgment basis generation unit 15.

[0023] When the output unit 14 determines not to display the judgment basis information (when the difficulty determination unit 13 determines that recognition is easy), the judgment basis generation unit 15 does not generate the judgment basis information. On the other hand, when the output unit 14 determines that the judgment basis information is to be displayed (when the difficulty determination unit 13 determines that recognition is difficult), the judgment basis generation unit 15 generates the judgment basis information and outputs it to the display unit 16.

[0024] Specifically, when the output unit 14 determines that the determination basis information should be displayed, the determination basis generation unit 15 determines the characteristics of the vessel pattern and the surface pattern based on the endoscopic image input from the endoscopic image input unit 11, and generates the determination basis information. Note that the determination basis generation unit 15 may generate an image description for explaining the input endoscopic image by AI captioning.

[0025] When judgment basis information has not been generated by the judgment basis generation unit 15, the display unit 16 displays the endoscopic image input from the endoscopic image input unit 11 and the recognition result input from the recognition result receiving unit 12.

[0026] On the other hand, when judgment basis information is generated by the judgment basis generation unit 15, the display unit 16 displays the endoscopic image input from the endoscopic image input unit 11, the recognition result input from the recognition result receiving unit 12, and the judgment basis information generated by the judgment basis generation unit 15.

[0027] Fig. 2 is a diagram showing an example of a display screen displayed on the display unit when recognition is easy, and Fig. 3 is a diagram showing an example of a display screen displayed on the display unit when recognition is difficult.

[0028] As described above, when the difficulty determination unit 13 determines that recognition is easy, no determination basis information is generated by the determination basis generation unit 15. Therefore, as shown in Fig. 2, the endoscopic image 30 input from the endoscopic image input unit 11 and the recognition result 31 input from the recognition result receiving unit 12 are displayed on the display unit 16.

[0029] On the other hand, if the difficulty determination unit 13 determines that recognition is difficult, the determination basis generation unit 15 generates determination basis information and outputs it to the display unit 16. Therefore, as shown in FIG. 3 , the endoscopic image 30 input from the endoscopic image input unit 11, the recognition result 31 input from the recognition result receiving unit 12, and the determination basis information 32 input from the determination basis generation unit 15 are displayed on the display unit 16. Note that the present invention can also be used in conjunction with CADe (Computer-Aided Detection) for detecting the position of the lesion 50. When CADe is used in conjunction with the present invention, the detected position of the lesion may be indicated by a marker such as a bounding box, as exemplified by reference numeral 40 in FIGS. 2 and 3 .

[0030] 4 is a flowchart showing an example of the operation of the endoscopic image processing device of the first embodiment. First, an endoscopic image is input to the endoscopic image input unit 11 (S1). Next, a recognition result in which a lesion is recognized from the endoscopic image is received (S2). This recognition result is generated by the recognition result generation unit 17 and received by the recognition result receiving unit 12.

[0031] Next, the difficulty determination unit 13 determines whether or not it is difficult to recognize the lesion based on the recognition result (S3). If it is determined that it is difficult to recognize the lesion (S3: YES), the determination basis generation unit 15 generates determination basis information (S4). Next, the endoscopic image, the recognition result, and the determination basis information are displayed on the display unit 16 (S5), and the process ends.

[0032] On the other hand, if it is determined that the lesion is not difficult (easy) to recognize (S3: NO), the endoscopic image and the recognition result are displayed on the display unit 16 (S6), and the process ends. In other words, if the lesion is easy to recognize, the determination basis information is not generated, and only the endoscopic image and the recognition result are displayed on the display unit 16.

[0033] As described above, in the endoscopic image processing device 1 of this embodiment, when it is easy to recognize (determine) a lesion, the basis information is not displayed on the display unit 16. On the other hand, in the endoscopic image processing device 1, only when it is difficult to recognize (determine) a lesion is the basis information displayed on the display unit 16. In this way, the endoscopic image processing device 1 of this embodiment can appropriately display the determination basis information on the display unit 16 based on whether it is difficult to recognize a lesion.

[0034] (Modification 1) Fig. 5 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 1 of the first embodiment. In Fig. 5, the same components as those in Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0035] The endoscopic image processing device 1A includes a difficulty determination unit 13A instead of the difficulty determination unit 13 in Fig. 1. The endoscopic image input to the endoscopic image input unit 11 is input to the difficulty determination unit 13A.

[0036] The difficulty determination unit 13A determines whether recognition is difficult based on the features of the endoscopic image input from the endoscopic image input unit 11. Specifically, the difficulty determination unit 13A calculates the feature amounts of the lesion shown in the endoscopic image (for example, the edge amount, the color difference from the surrounding mucosa, the degree of irregularity of the vascular manipulation pattern, the degree of mixing of regions of different classifications, etc.), and determines that the judgment is difficult if the calculated feature amount is equal to or greater than a certain value.

[0037] The other configurations are the same as those of the first embodiment. The output unit 14 determines whether or not to display the basis information based on the determination result of the difficulty determination unit 13A. If the difficulty determination unit 13A determines that the determination is difficult, the output unit causes the determination basis generation unit 15 to generate basis information and causes the generated basis information to be displayed on the display unit 16.

[0038] (Modification 2) Fig. 6 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 2 of the first embodiment. In Fig. 6, the same components as those in Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0039] The endoscopic image processing device 1B includes a difficulty determination unit 13B instead of the difficulty determination unit 13 in Fig. 1. The doctor's determination result inputted via the doctor determination information input unit 18 is inputted to the difficulty determination unit 13B.

[0040] The doctor's judgment information input unit 18 is, for example, a microphone for voice input, a keyboard for character input, a foot pedal operated by the user, etc. The doctor inputs the lesion judgment result using the doctor's judgment information input unit 18. The doctor's judgment result input by the doctor's judgment information input unit 18 is input to the difficulty judgment unit 13B.

[0041] The difficulty determination unit 13B determines whether recognition is difficult by comparing the recognition result received by the recognition result receiving unit 12 with the doctor's judgment result input via the doctor judgment information input unit 18. Specifically, the difficulty determination unit 13B determines that recognition is difficult when the judgment result of the recognition result generating unit 17 differs from the doctor's judgment result.

[0042] The other configurations are the same as those of the first embodiment. The output unit 14 determines whether or not to display the basis information based on the determination result of the difficulty determination unit 13B. When the difficulty determination unit 13B determines that recognition is difficult, the output unit 14 causes the determination basis generation unit 15 to generate basis information and causes the generated basis information to be displayed on the display unit 16.

[0043] (Modification 3) Fig. 7 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 3 of the first embodiment. In Fig. 7, the same components as those in Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0044] The endoscopic image processing device 1C includes a difficulty determination unit 13C instead of the difficulty determination unit 13 in Fig. 1. Imaging information 19 is input to the difficulty determination unit 13C. The imaging information 19 includes any one of imaging region information, imaging condition information, and imaging model information.

[0045] The difficulty determination unit 13C determines that recognition is difficult when the imaging site information indicates that the imaging site of the image is a predetermined organ or a predetermined site included in the predetermined organ. Specifically, the predetermined organ is, for example, the stomach or a stomach with a history of Helicobacter pylori infection. Furthermore, the predetermined site is, for example, the cardia of the stomach, the submucosa of the large intestine, or cancer in intestinal metaplasia.

[0046] The imaging condition information is light information (observation light information) when the endoscopic image is captured. The difficulty determination unit 13C receives the light information (observation light information) when the endoscopic image is captured, and determines that recognition is difficult if the light when the endoscopic image is captured is white light.

[0047] Furthermore, the difficulty determination unit 13C determines that recognition is difficult when the imaging part is a predetermined organ or an imaging part included in a predetermined organ, based on the imaging model information.

[0048] The other configurations are the same as those of the first embodiment. The output unit 14 determines whether or not to display the basis information based on the determination result of the difficulty determination unit 13C. When the difficulty determination unit 13C determines that recognition is difficult, the output unit 14 causes the determination basis generation unit 15 to generate basis information and causes the generated basis information to be displayed on the display unit 16.

[0049] (Modification 4) Fig. 8 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 4 of the first embodiment. In Fig. 8, the same components as those in Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0050] The endoscopic image processing device 1D includes a difficulty determination unit 13D instead of the difficulty determination unit 13 in Fig. 1. Physician proficiency information 20 indicating the proficiency of a physician is input to the difficulty determination unit 13D. The physician proficiency information 20 includes, for example, years of experience, number of cases experienced, qualifications, etc. This physician proficiency information 20 is input in advance before an endoscopic procedure is performed.

[0051] The difficulty determination unit 13D calculates the feature amount of the lesion shown in the image, and determines that the judgment is difficult if the feature amount is equal to or greater than a predetermined threshold. At this time, the difficulty determination unit 13D changes the predetermined threshold based on the doctor proficiency information 20. For example, the predetermined threshold is set to be higher as the doctor's proficiency level based on the doctor proficiency information 20 increases.

[0052] The other configurations are the same as those of the first embodiment. The output unit 14 determines whether or not to display the basis information based on the determination result of the difficulty determination unit 13D. When the difficulty determination unit 13D determines that recognition is difficult, the output unit 14 causes the determination basis generation unit 15 to generate basis information and causes the generated basis information to be displayed on the display unit 16.

[0053] As a result, even if a certain lesion is being observed, if an experienced doctor is observing, the basis for the judgment is not displayed on the display unit 16, but if an unskilled doctor is observing, the basis information is displayed on the display unit 16.

[0054] Second Embodiment Next, a second embodiment will be described. Fig. 9 is a block diagram showing the configuration of an endoscopic image processing apparatus according to the second embodiment. In Fig. 9, the same components as those in Fig. 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0055] As shown in FIG. 1, the endoscopic image processing apparatus 1E includes an output unit 14A and a determination basis generating unit 15A instead of the output unit 14 and the determination basis generating unit 15 of FIG.

[0056] The determination basis generating unit 15A generates determination basis information based on the endoscopic image input from the endoscopic image input unit 11, and outputs the generated information to the output unit 14A.

[0057] When the difficulty determination unit 13 determines that it is difficult to recognize the lesion, the output unit 14A inputs the determination basis information generated by the determination basis generation unit 15A to the display unit 16, and when the difficulty determination unit 13 determines that it is not difficult to recognize the lesion, the output unit 14A blocks the generated determination basis information from being input to the display unit 16. In other words, when the difficulty determination unit 13 determines that it is not difficult to recognize the lesion, the output unit 14A maintains the display content of the display unit 16, on which the endoscopic image and the recognition result are displayed.

[0058] In the first embodiment, when the difficulty determining unit 13 determines that it is difficult to recognize a lesion, the determination basis generating unit 15 generates determination basis information and displays it on the display unit 16 .

[0059] In contrast, in the second embodiment, the judgment basis generating unit 15A constantly generates judgment basis information, and when the difficulty determining unit 13 determines that it is difficult to recognize a lesion, the output unit 14A controls the display unit 16 to display the generated judgment basis information. The other configurations are the same as those of the first embodiment.

[0060] 10 is a flowchart showing an example of the operation of the endoscopic image processing device of the second embodiment. First, an endoscopic image is input to the endoscopic image input unit 11 (S11). Next, a recognition result in which a lesion is recognized from the endoscopic image is received (S12). This recognition result is generated by the recognition result generation unit 17 and received by the recognition result receiving unit 12. Next, determination basis information is generated by the determination basis generation unit 15A (S13).

[0061] Next, the difficulty determination unit 13 determines whether or not it is difficult to recognize the lesion based on the recognition result (S14). If it is determined that it is difficult to recognize the lesion (S14: YES), the endoscopic image, the recognition result, and the determination basis information are displayed on the display unit 16 (S15), and the process ends.

[0062] On the other hand, if it is determined that the lesion recognition is not difficult (easy) (S14: NO), the output unit 14 blocks the output of the generated determination basis information (S16), and the endoscopic image and the recognition result are displayed on the display unit 16 (S17), and the process ends.

[0063] Through the above processing, the endoscopic image processing device 1E of this embodiment, like the first embodiment, can appropriately display determination basis information on the display unit 16 based on whether or not it is difficult to recognize a lesion.

[0064] (Modification 1) Fig. 11 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 1 of the second embodiment. In Fig. 11, the same components as those in Fig. 9 are denoted by the same reference numerals, and the description thereof will be omitted.

[0065] The endoscopic image processing device 1F includes a difficulty determining unit 13E instead of the difficulty determining unit 13 in Fig. 9. The difficulty determining unit 13E receives determination basis information from the determination basis generating unit 15A.

[0066] The difficulty determination unit 13E determines the difficulty of recognizing a lesion from the determination basis information input from the determination basis generation unit 15A, and outputs the determination result to the output unit 14A. Specifically, the difficulty determination unit 13E determines that it is difficult to recognize a lesion when the determination basis information includes both the basis for not being able to determine that a lesion exists and the basis for determining that a lesion exists.

[0067] The basis for determining that an area is not a lesion is information such as the absence of protrusion, the absence of color difference from the surrounding mucosa, the absence of capillary aggregation, the absence of difference in texture from the surrounding mucosa, or the absence of bleeding.On the other hand, the basis for determining that an area is a lesion is information such as the presence of protrusion, the presence of color difference from the surrounding mucosa, the presence of capillary aggregation, the presence of difference in texture from the surrounding mucosa, or the presence of bleeding.

[0068] The other configurations are the same as those of the second embodiment. The output unit 14A controls whether to block the generated determination basis information or output it to the display unit 16, based on the determination result of the difficulty determination unit 13E.

[0069] (Modification 2) Fig. 12 is a block diagram showing the configuration of an endoscopic image processing device according to Modification 2 of the second embodiment. In Fig. 12, the same components as those in Fig. 9 are denoted by the same reference numerals and description thereof will be omitted.

[0070] The endoscopic image processing device 1G includes a difficulty determination unit 13F instead of the difficulty determination unit 13 in FIG.

[0071] The difficulty determination unit 13F may determine the difficulty of recognizing a lesion based on the endoscopic image input from the endoscopic image input unit 11, similar to the first modification of the first embodiment.

[0072] Furthermore, the difficulty determination unit 13F may determine the difficulty of recognizing a lesion based on the doctor's judgment result input from the doctor judgment information input unit 18, as in the second modification of the first embodiment.

[0073] Furthermore, the difficulty determination unit 13F may determine the difficulty of recognizing a lesion based on the imaging information 19, similar to the third modification of the first embodiment.

[0074] Furthermore, the difficulty determination unit 13F may determine the difficulty of recognizing a lesion based on doctor proficiency information 20 indicating the doctor's proficiency, similar to the fourth modification of the first embodiment.

[0075] The difficulty determination unit 13F may also determine the difficulty of recognizing a lesion by determining the movement of the endoscope. For example, the movement of the endoscope can be determined by calculating the amount of movement based on the optical flow of the endoscopic image.

[0076] If the amount of movement of the endoscope is below a certain value, it is highly likely that the doctor is continuing to observe the same lesion, and it is conceivable that the doctor is unsure of his or her judgment. Therefore, if the amount of movement of the endoscope is below a certain value, the difficulty determination unit 13F determines that it is difficult to recognize the lesion.

[0077] The other configurations are the same as those of the second embodiment. The output unit 14A controls whether to block the generated determination basis information or output it to the display unit 16, based on the determination result of the difficulty determination unit 13F.

[0078] The endoscopic image processing device 1 described in the above-mentioned embodiment or modified example can be used together with an endoscope 30 and a recognition result generation unit (recognition result generation device) 17 connected to the endoscope 30, as illustrated in FIG. 13.

[0079] Fig. 13 is a diagram showing the configuration of an endoscope system. As shown in Fig. 13, an endoscopic image processing device 1 is connected to a recognition result generating device and an endoscope 30. In addition, the recognition result generating device is connected to the endoscope 30.

[0080] The endoscopic image processing device 1 may be connected to the recognition result generating device and the endoscope 30 by wire or wirelessly. The recognition result generating device may be connected to the endoscope 30 by wire or wirelessly.

[0081] It should be noted that the steps in the flowcharts in this specification may be executed in a different order, may be executed multiple times simultaneously, or may be executed in a different order each time, as long as this does not contradict the nature of the steps.

[0082] The present invention is not limited to the above-described embodiments, and various modifications, combinations, and applications are possible without departing from the spirit of the invention. For example, the present invention can be combined with a technology for turning on / off functions related to XAI. In this case, the XAI itself may be turned on / off, or the XAI itself may be left on while the display adjustment function for XAI based on the difficulty of recognition may be turned on / off.

[0083] Furthermore, among the technologies described herein, many of the controls and functions, mainly those described in the flowcharts, can be set by a program, and the above-described controls and functions can be realized by having a computer read and execute the program. The program can be recorded or stored, in whole or in part, as a computer program product on a portable medium such as a flexible disk, CD-ROM, or nonvolatile memory, or on a storage medium such as a hard disk or volatile memory, and can be distributed or provided at the time of product shipment, via a portable medium, or via a communication line. A user can easily realize the endoscopic image processing program of this embodiment by downloading the program via a communication network and installing it on a computer, or by installing it on a computer from a recording medium.

Claims

1. An endoscopic image processing device including: a recognition receiving unit that receives recognition results of lesions recognized from images; a basis information generating unit that generates basis information for the recognition results; a difficulty determining unit that determines the difficulty of recognition based on recognition difficulty determination materials; and an output unit that outputs the basis information to a display unit when the difficulty determining unit determines that recognition is difficult.

2. An endoscopic image processing device as described in claim 1, wherein the recognition difficulty determination material is at least one of the recognition result, the image, the basis information, the imaging area of ​​the image, the imaging conditions of the image, or the model of the imaging device that captured the image.

3. The endoscopic image processing device described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition from the recognition result, and if the difficulty determination unit determines that recognition is difficult, the output unit inputs the recognition result to the basis information generation unit to generate the basis information, and outputs the generated basis information to the display unit.

4. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition during or after the generation of the basis information, and when the difficulty determination unit determines that recognition is not difficult, the output unit blocks the generated basis information from being input to the display unit.

5. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition from the recognition result, and the difficulty determination unit determines that recognition is difficult if the certainty of the recognition result is less than a predetermined value.

6. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the recognition difficulty from the recognition result, and the difficulty determination unit determines that recognition is difficult when the recognition result is a predetermined disease name.

7. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition from the image, and the difficulty determination unit determines that recognition is difficult if the color difference between the lesion and its surroundings in the image is less than a threshold value.

8. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the recognition difficulty from the image, and the difficulty determination unit determines that recognition is difficult when the edge of the lesion in the image is less than a threshold value.

9. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit receives light information when the image is captured, the difficulty determination unit determines the difficulty of recognition from the image, and the difficulty determination unit determines that recognition is difficult if the light when the image is captured is white light.

10. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition from the basis information, and the difficulty determination unit determines that recognition is difficult when the basis information includes both basis for not determining that there is a lesion and basis for determining that there is a lesion.

11. An endoscopic image processing device as described in claim 2, wherein the difficulty determination unit determines the difficulty of recognition from the imaging area of ​​the image, and the difficulty determination unit determines that recognition is difficult if the imaging area of ​​the image is a specified organ or a specified area contained in a specified organ.

12. The endoscopic image processing device according to claim 1, wherein the difficulty determination unit is capable of changing the threshold value of the recognition difficulty.

13. An endoscopic image processing device as described in claim 12, wherein the difficulty determination unit includes a proficiency information input unit for inputting proficiency information, and the recognition difficulty threshold is changeable according to the input proficiency information.

14. An endoscopic image processing device includes a processor that receives a recognition result of a lesion recognized from an image, generates basis information for the recognition result, determines the difficulty of recognition based on at least one of the recognition result, the image, the basis information, or the imaged area of ​​the image, and outputs the basis information to a monitor if it determines that recognition is difficult.

15. An endoscopic image processing method that receives a recognition result of a lesion recognized from an image, generates basis information for the recognition result, determines the difficulty of recognition, and outputs the basis information to a monitor if it is determined that recognition is difficult.

16. An endoscopic image processing program that causes a computer to execute the following processes: receiving a recognition result of recognizing a lesion from an image; generating basis information for the recognition result; determining the difficulty of recognition; and outputting the basis information to a monitor if it is determined that recognition is difficult.

17. A non-volatile storage medium storing an endoscopic image processing program, which causes a computer to execute the following processes: receiving a recognition result of recognizing a lesion from an image; generating basis information for the recognition result; determining the difficulty of recognition; and outputting the basis information to a monitor if it is determined that recognition is difficult.

18. An endoscope system comprising: an endoscope; a recognition result generating device connected to the endoscope; and an endoscopic image processing device according to claim 1 connected to the endoscope and the recognition result generating device.

Images