阅读说明：本技术 诊断辅助装置、诊断辅助方法以及程序 (Diagnosis support device, diagnosis support method, and program ) 是由 山田达喜 窪田明广 神田大和 木村光隆 北村诚 于 2019-03-18 设计创作，主要内容包括：诊断辅助装置具有：异常症状确定部,其构成为进行如下处理：根据身体信息以及内窥镜图像中的至少一方,确定在诊断对象脏器中出现的一个异常症状,其中身体信息包含能够估计被检者的诊断对象脏器的状态的1个以上的信息,内窥镜图像是拍摄诊断对象脏器而得到的图像；病变提取功能部,其构成为具有多个病变提取部,多个病变提取部进行针对在诊断对象脏器中可能出现的每个异常症状进行了特殊化的不同的处理,作为用于从内窥镜图像中提取病变候选区域的病变提取处理；以及病变提取功能控制部,其构成为进行用于从多个病变提取部中选择一个病变提取部的处理,并且对病变提取功能部进行用于在一个病变提取部中进行病变提取处理的控制,其中一个病变提取部与由异常症状确定部确定出的一个异常症状对应。(The diagnosis assistance device includes: an abnormal symptom identification unit configured to perform: specifying an abnormal symptom appearing in a target organ to be diagnosed based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information capable of estimating a state of the target organ to be diagnosed of a subject, the endoscopic image being an image obtained by imaging the target organ to be diagnosed; a lesion extraction function unit configured to include a plurality of lesion extraction units that perform different processes specialized for each abnormal symptom that may appear in a diagnosis target organ as lesion extraction processes for extracting a lesion candidate region from an endoscopic image; and a lesion extraction function control unit configured to perform processing for selecting one lesion extraction unit from the plurality of lesion extraction units, and to perform control for performing lesion extraction processing in one lesion extraction unit, one of the lesion extraction units corresponding to the one abnormal symptom specified by the abnormal symptom specifying unit.)

1. A diagnosis assistance apparatus characterized by comprising:

an abnormal symptom identification unit configured to perform: a function of specifying an abnormal symptom occurring in a target organ for diagnosis based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information that can estimate a state of the target organ for diagnosis of a subject, the endoscopic image being an image obtained by imaging the target organ for diagnosis;

a lesion extraction function unit configured to include a plurality of lesion extraction units that perform different processes specialized for each abnormal symptom that may appear in the organ to be diagnosed as a lesion extraction process for extracting a lesion candidate region from the endoscopic image; and

a lesion extraction function control unit configured to perform processing for selecting one lesion extraction unit from the plurality of lesion extraction units, the one lesion extraction unit corresponding to the one abnormal symptom specified by the abnormal symptom specifying unit, and to perform control for performing the lesion extraction processing in the one lesion extraction unit with respect to the lesion extraction function unit.

2. The diagnostic aid of claim 1,

at least one of the following information is contained in the body information: information indicating a test result obtained by performing a test on the organ to be diagnosed, information indicating a treatment history of the organ to be diagnosed, information indicating an administration history of a drug related to the organ to be diagnosed, and information indicating the presence or absence of a self-perceived symptom of the subject.

3. The diagnostic aid of claim 1,

the abnormal symptom specifying unit is configured to further perform processing for specifying a magnitude of a lesion risk in the one abnormal symptom.

4. The diagnostic aid of claim 3,

the lesion extraction function control unit is configured to perform processing for setting a sensitivity corresponding to the size of the lesion risk determined by the abnormal symptom determination unit, and to perform control for performing the lesion extraction processing based on the sensitivity in the one lesion extraction unit.

5. The diagnostic aid of claim 1,

the abnormal symptom determination unit is configured to further perform a process of determining whether or not to recommend special light observation for the one abnormal symptom.

6. The diagnostic aid of claim 1,

the lesion extraction function unit is configured to further include another lesion extraction unit that performs, as the lesion extraction process, a process that is not specialized with respect to a specific abnormal symptom that may occur in the organ to be diagnosed,

the lesion extraction function control unit is configured to perform processing for selecting the other lesion extraction unit and to perform control for performing the lesion extraction processing in the other lesion extraction unit, when the one abnormal symptom is not specified by the abnormal symptom specifying unit.

7. The diagnostic aid of claim 1,

the abnormal symptom specifying unit is configured to further perform a process of specifying a site to be observed corresponding to an anatomical region of the organ to be diagnosed included in the endoscopic image, and a process of specifying a size of a lesion risk corresponding to the one abnormal symptom appearing in the site to be observed.

8. The diagnostic aid of claim 7,

the lesion extraction function control unit is configured to perform processing for setting a sensitivity corresponding to the size of the lesion risk determined by the abnormal symptom determination unit, and to perform control for performing the lesion extraction processing based on the sensitivity in the one lesion extraction unit.

9. The diagnostic aid of claim 1,

the diagnosis support device further includes a display control unit configured to perform processing for displaying the endoscopic image and information indicating the position of the lesion candidate region extracted by the one lesion extraction unit on a display device together.

10. The diagnostic aid of claim 5,

the diagnosis support apparatus further includes a display control unit configured to perform processing for displaying the endoscopic image, information indicating the position of the lesion candidate region extracted by the one lesion extraction unit, and information prompting execution of the special light observation on a display device together when it is detected that the recommendation for the one abnormal symptom to be switched to the special light observation is made.

11. The diagnostic aid of claim 1,

the lesion extraction function control unit is configured to perform a process for setting a sensitivity in accordance with an instruction from a user, and to perform a control for performing the lesion extraction process based on the sensitivity in the one lesion extraction unit.

12. A diagnosis assistance method characterized by comprising the steps of,

the abnormal symptom determination unit performs the following processing: for specifying an abnormal symptom appearing in a target organ for diagnosis based on at least one of body information including 1 or more pieces of information capable of estimating a state of the target organ for diagnosis of a subject and an endoscopic image obtained by imaging the target organ for diagnosis,

a plurality of lesion extraction units in the lesion extraction function unit perform different processes specialized for each abnormal symptom that may appear in the diagnosis target organ, as lesion extraction processes for extracting a lesion candidate region from the endoscopic image,

the lesion extraction function control unit performs processing for selecting one lesion extraction unit corresponding to the one abnormal symptom determined by the abnormal symptom determination unit from the plurality of lesion extraction units, and performs control for performing the lesion extraction processing in the one lesion extraction unit on the lesion extraction function unit.

13. A program for causing a computer to execute:

a function of specifying an abnormal symptom occurring in a target organ for diagnosis based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information that can estimate a state of the target organ for diagnosis of a subject, the endoscopic image being an image obtained by imaging the target organ for diagnosis;

selecting one lesion extraction unit corresponding to the one abnormal symptom from a plurality of lesion extraction units that perform different processes specialized for each abnormal symptom that may appear in the diagnosis target organ as lesion extraction processes for extracting a lesion candidate region from the endoscopic image; and

the lesion extraction processing is performed in the one lesion extraction unit.

Technical Field

The invention relates to a diagnosis support device, a diagnosis support method, and a program.

Background

In endoscopy in the medical field, as a technique for presenting Diagnosis support information for diagnosing a lesion present in an image obtained by imaging a subject such as a living tissue to a user such as a doctor, for example, Computer-Aided Diagnosis (CAD) is known. Further, for example, japanese patent application laid-open No. 2011-2120094 proposes a system configured by using a CAD technique.

Specifically, japanese patent application laid-open publication No. 2011-2120094 discloses the following structure: in a system including a plurality of lesion extraction units capable of extracting a plurality of lesions of different types, diagnosis information is acquired based on patient information, one lesion extraction unit corresponding to a determination result of the presence or absence of a definitive diagnosis in the diagnosis information is selected from the plurality of lesion extraction units, and the lesion extracted by the one lesion extraction unit is displayed.

However, japanese patent application laid-open No. 2011-212094 does not disclose a method for selecting one lesion extraction unit from a plurality of lesion extraction units, in particular, while considering factors that affect the accuracy of lesion extraction in a target organ for diagnosis. Therefore, according to the structure disclosed in japanese patent application laid-open No. 2011-2120094, for example, the following problems exist: when a factor that hinders detection of a lesion occurs in a target organ to be diagnosed due to the influence of histology corresponding to a medical history of a subject or the like, accuracy of extraction of the lesion in the target organ to be diagnosed is lowered.

The present invention has been made in view of the above circumstances, and an object thereof is to provide a diagnosis support device, a diagnosis support method, and a program that can ensure accuracy of extracting a lesion in a target organ for diagnosis.

Disclosure of Invention

Means for solving the problems

A diagnosis assistance device according to an aspect of the present invention includes: an abnormal symptom identification unit configured to perform: specifying an abnormal symptom appearing in a diagnosis target organ based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information that can estimate a state of the diagnosis target organ of a subject, the endoscopic image being an image obtained by imaging the diagnosis target organ; a lesion extraction function unit configured to include a plurality of lesion extraction units that perform different processes specialized for each abnormal symptom that may appear in the organ to be diagnosed as a lesion extraction process for extracting a lesion candidate region from the endoscopic image; and a lesion extraction function control unit configured to perform processing for selecting one lesion extraction unit from the plurality of lesion extraction units, the one lesion extraction unit corresponding to the one abnormal symptom specified by the abnormal symptom specifying unit, and to perform control for performing the lesion extraction processing in the one lesion extraction unit with respect to the lesion extraction function unit.

In a diagnosis support method according to an aspect of the present invention, an abnormal symptom specifying unit performs: specifying one abnormal symptom appearing in a diagnosis target organ based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information capable of estimating a state of the diagnosis target organ of a subject, the endoscopic image being an image obtained by imaging the diagnosis target organ, a plurality of lesion extracting units in a lesion extracting function unit performing different processing specialized for each abnormal symptom that may appear in the diagnosis target organ as lesion extracting processing for extracting a lesion candidate region from the endoscopic image, a lesion extracting function control unit performing processing for selecting one lesion extracting unit from the plurality of lesion extracting units and performing control for performing the lesion extracting processing in the one lesion extracting unit on the lesion extracting function unit, wherein the one lesion extracting section corresponds to the one abnormal symptom determined by the abnormal symptom determining section.

A program according to an aspect of the present invention causes a computer to execute: specifying an abnormal symptom appearing in a diagnosis target organ based on at least one of body information and an endoscopic image, the body information including 1 or more pieces of information that can estimate a state of the diagnosis target organ of a subject, the endoscopic image being an image obtained by imaging the diagnosis target organ; selecting one lesion extraction unit corresponding to the one abnormal symptom from a plurality of lesion extraction units that perform different processes specialized for each abnormal symptom that may appear in the diagnosis target organ as lesion extraction processes for extracting a lesion candidate region from the endoscopic image; and performing the lesion extraction process in the one lesion extraction unit.

Drawings

Fig. 1 is a diagram showing a configuration of a main part of an endoscope system including a diagnosis support device according to an embodiment.

Fig. 2 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 1.

Fig. 3 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to embodiment 1.

Fig. 4 is a diagram showing an example of a display image displayed by the processing of the diagnosis assistance apparatus according to embodiment 1.

Fig. 5 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to the modification example of embodiment 1.

Fig. 6 is a diagram showing an example of a display image displayed by processing in the diagnosis assistance apparatus according to the modification example of embodiment 1.

Fig. 7 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 2.

Fig. 8 is a diagram showing an example of a display image displayed by the processing of the diagnosis support apparatus according to embodiment 2.

Fig. 9 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 3.

Fig. 10 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to embodiment 3.

Detailed Description

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(embodiment 1)

Fig. 1 to 6 are views of embodiment 1.

As shown in fig. 1, the endoscope system 1 is configured to include an endoscope 11, a main body device 12, a diagnosis support device 13, and a display device 14. Fig. 1 is a diagram showing a configuration of a main part of an endoscope system including a diagnosis support device according to an embodiment.

The endoscope 11 includes, for example, an elongated insertion portion (not shown) that can be inserted into a subject, and an operation portion (not shown) provided at a proximal end portion of the insertion portion. The endoscope 11 is configured to be detachably connected to the main body apparatus 12 via a universal cable (not shown) extending from the operation unit, for example. Further, a light guide member (not shown) such as an optical fiber for guiding illumination light supplied from the main body device 12 and emitting the illumination light from the distal end portion of the insertion portion is provided inside the endoscope 11. Further, an imaging unit 111 is provided at the distal end of the insertion portion of the endoscope 11.

The imaging unit 111 is configured to include an imaging element such as a CCD image sensor or a CMOS image sensor, for example. The image pickup unit 111 is configured to pick up an image of return light from an object illuminated with illumination light emitted through the distal end portion of the insertion portion of the endoscope 11, generate an image pickup signal corresponding to the picked-up return light, and output the image pickup signal to the main body apparatus 12.

The main body device 12 is configured to be detachably connected to the endoscope 11 and the diagnosis assisting device 13, respectively. As shown in fig. 1, the main body device 12 is configured to include, for example, a light source unit 121, an image generating unit 122, an input I/F (interface) 123, a control unit 124, and a storage medium 125.

The light source unit 121 includes 1 or more light emitting elements such as LEDs, for example. Specifically, the light source unit 121 includes, for example, a blue LED that generates blue light (hereinafter, also referred to as B light), a green LED that generates green light (hereinafter, also referred to as G light), and a red LED that generates red light (hereinafter, also referred to as R light). The light source unit 121 is configured to be able to generate BN light corresponding to narrow-band light having a peak wavelength near 415nm and GN light corresponding to narrow-band light having a peak wavelength near 540nm by restricting the wavelength bands of B light and G light by an optical filter or the like, for example. The light source unit 121 is configured to generate illumination light according to the control of the control unit 124 and supply the illumination light to the endoscope 11.

The image generation unit 122 is configured to generate an endoscopic image based on the image pickup signal output from the endoscope 11, and sequentially output the generated endoscopic image to the diagnosis assisting apparatus 13 in 1 frame and 1 frame.

The input I/F123 is configured to include 1 or more switches and/or buttons capable of inputting instructions, information, and the like corresponding to user operations. Specifically, the input I/F123 is configured to be able to input, for example, body information including 1 or more pieces of information, the 1 or more pieces of information being able to estimate the state of a diagnostic organ of a subject under endoscopic examination. The input I/F123 is provided with an observation mode changeover switch that can set an observation mode for performing observation by the endoscope 11 to either a white light observation mode for observing an object irradiated with B light, G light, and R light as illumination light or a special light observation (narrow band observation) mode for observing an object irradiated with BN light and GN light as illumination light. Further, the input I/F123 is provided with an abnormality finding detection instruction switch that can instruct detection of abnormality finding from the endoscope image currently under observation (hereinafter, simply referred to as an abnormality finding detection instruction). The observation mode changeover switch and/or the abnormality detection instruction switch are not limited to the input I/F123 provided in the main body device 12, and may be provided in, for example, an operation portion of the endoscope 11.

The control unit 124 performs control related to the operations of the endoscope 11 and the main body apparatus 12 based on the instruction input by the input I/F123. The control unit 124 is configured to be capable of performing an operation for outputting an instruction input by the input I/F123 to the diagnosis assisting apparatus 13. The control unit 124 is configured to be capable of performing an operation for outputting information such as the physical information input by the input I/F123 to the diagnosis assisting apparatus 13.

In the present embodiment, the image generation unit 122 and the control unit 124 of the main body device 12 may be configured as respective electronic circuits, or may be configured as circuit modules in an integrated circuit such as an FPGA (Field Programmable Gate Array). In the present embodiment, for example, the main device 12 may be configured to include 1 or more processors (CPUs and the like). Further, by appropriately changing the configuration of the present embodiment, for example, the computer may be caused to read a program for executing the functions of the image generation unit 122 and the control unit 124 from the storage medium 125 such as a memory, and to perform an operation corresponding to the read program.

The diagnosis assistance device 13 is configured as a computer having 1 or more processors 131 and storage media 132, for example. The diagnosis support device 13 is configured to be detachably connected to the main body device 12 and the display device 14, respectively. The diagnosis assisting apparatus 13 is configured to perform lesion extraction processing for extracting a lesion candidate region from an endoscopic image output from the main apparatus 12, based on the instruction and/or information output from the main apparatus 12. The lesion candidate region is extracted as a region in which an abnormality such as a benign tumor or a malignant tumor is found. The diagnosis assisting apparatus 13 is configured to generate a display image by adding visual information indicating the position of the lesion candidate region extracted by the above-described lesion extraction processing to the endoscopic image, and output the generated display image to the display device 14. As shown in fig. 2, for example, the diagnosis assisting apparatus 13 includes an abnormal symptom specifying unit 13A, a lesion extraction function control unit 13B, a lesion extraction function unit 13C, and a display control unit 13D. Fig. 2 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 1.

The abnormal symptom specifying unit 13A performs processing for specifying one abnormal symptom appearing in a diagnosis target organ of a subject under endoscopic examination based on the body information output from the main body apparatus 12. The abnormal symptom specifying unit 13A is configured to output information indicating one abnormal symptom specified by the above-described processing to the lesion extraction function control unit 13B.

The lesion extraction function control unit 13B is configured to perform processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, based on the information output from the abnormal symptom specifying unit 13A. The lesion extraction function control unit 13B is configured to perform control for performing a lesion extraction process in one lesion extraction unit selected by the above-described process.

The lesion extraction function unit 13C is configured to include a plurality of lesion extraction units LES and a lesion extraction unit LESY.

The lesion extraction units LES are configured to perform different processes specialized for each abnormal symptom that may occur in a target organ to be diagnosed of a subject under an endoscopic examination, as lesion extraction processing for extracting a lesion candidate region from an endoscopic image obtained by imaging the target organ to be diagnosed.

The lesion extraction unit LESY is configured to perform, as lesion extraction processing for extracting a lesion candidate region from an endoscopic image obtained by imaging a diagnostic organ of a subject under endoscopic examination, processing that is not specialized for a specific abnormal symptom that may appear in the diagnostic organ.

The lesion extraction function unit 13C performs lesion extraction processing corresponding to processing for extracting a lesion candidate region from an endoscopic image outputted from the main body apparatus 12, in any one of the lesion extraction unit LESX selected from the plurality of lesion extraction units LES or the lesion extraction unit LESY selected in place of the lesion extraction unit LESX, according to the control of the lesion extraction function control unit 13B. The lesion extraction function unit 13C is configured to output information capable of specifying the position of a lesion candidate region extracted by the lesion extraction process of the lesion extraction unit LESX or the lesion extraction unit LESY to the display control unit 13D.

The display control unit 13D is configured to perform the following processing based on the information output from the lesion extraction function unit 13C: the position of the lesion candidate region in the endoscopic image output from the main apparatus 12 is determined, and visual information indicating the position of the determined lesion candidate region is generated. The display control unit 13D is configured to perform the following processes: the above-described visual information is added to the endoscopic image to generate a display image, and the generated display image is output to the display device 14.

In the present embodiment, at least a part of the functions of the diagnosis assistance apparatus 13 may be implemented by the processor 131. In the present embodiment, at least a part of the diagnosis assistance device 13 may be configured as each electronic circuit, or may be configured as a circuit module in an integrated circuit such as an FPGA (Field Programmable Gate Array). Further, by appropriately changing the configuration of the present embodiment, for example, the computer may be caused to read a program for executing at least a part of the functions of the diagnosis support apparatus 13 from the storage medium 132 such as a memory, and to perform an operation corresponding to the read program.

The display device 14 is detachably connected to the diagnosis support device 13. The display device 14 includes a monitor and the like, and is configured to be capable of displaying a display image output from the diagnosis support device 13.

Next, the operation of the present embodiment will be described. In the following, a case where the organ to be diagnosed of the subject under endoscopic examination is the stomach will be described as an example. In the following, the observation mode when observation is performed with the endoscope 11 is set in advance as the white light observation mode.

After connecting the respective units of the endoscope system 1 and receiving power, a user such as an operator inputs 1 or more pieces of information capable of estimating the state of the stomach of the subject as body information of the subject under endoscopic examination by operating the input I/F123.

Specifically, as the physical information of the subject who receives the endoscopic examination, the user inputs information such as the examination result of Helicobacter pylori (Helicobacter pylori) examination, the examination result of pepsinogen examination, the presence or absence of the history of degerming treatment of Helicobacter pylori, the examination result of CDH1 gene, the presence or absence of the history of excision treatment, the examination result of HER2 gene, the examination result of EB virus examination, the presence or absence of NSAIDs (non-steroidal anti-inflammatory drugs) to be administered, and the presence or absence of self-perceived symptoms of the subject.

That is, according to the present embodiment, as the body information of the subject who receives the endoscopic examination, at least one of the information indicating the examination result obtained by performing the examination on the organ to be diagnosed, the information indicating the treatment history of the organ to be diagnosed, the information indicating the administration history of the drug related to the organ to be diagnosed, and the information indicating the presence or absence of the self-perceived symptom of the subject is input.

When the user inputs the body information of the subject to be examined under the endoscopic examination, the insertion portion of the endoscope 11 is inserted into the subject, and the distal end portion of the insertion portion is arranged at a position where a desired object in the stomach of the subject can be imaged.

When detecting that an instruction for completing the input of the body information of the subject who receives the endoscopic examination is given in the input I/F123, the control unit 124 performs an operation for outputting the body information to the diagnosis assisting apparatus 13.

When the power of the main body apparatus 12 is turned on, the control unit 124 controls the light source unit 121 to generate B light, G light, and R light as illumination light sequentially or simultaneously. Then, under the control of the control unit 124, illumination light is supplied from the light source unit 121 to the endoscope 11, return light from the subject illuminated with the illumination light is captured by the imaging unit 111, an endoscopic image EG corresponding to the imaging signal output from the imaging unit 111 to the main body apparatus 12 is generated by the image generation unit 122, and the generated endoscopic image EG is sequentially output to the diagnosis assistance apparatus 13 frame by frame.

The abnormal symptom specifying unit 13A performs processing for specifying one abnormal symptom appearing in the stomach of the subject under endoscopic examination based on the body information output from the main body apparatus 12. The abnormal symptom specifying unit 13A outputs information indicating one abnormal symptom specified by the above-described processing to the lesion extraction function control unit 13B.

Specifically, the abnormal symptom identification unit 13A reads the table data TDA shown in fig. 3 from the storage medium 132, and identifies one abnormal symptom appearing in the stomach of the subject under the endoscopic examination by comparing the read table data TDA with the body information output from the main body apparatus 12. Fig. 3 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to embodiment 1.

The "body information" column of the table data TDA in fig. 3 includes a plurality of items corresponding to body information (information input in the input I/F123) output from the main apparatus 12. In the "abnormal symptom" column of the table data TDA in fig. 3, a plurality of items indicating the symptoms of the stomach corresponding to the respective items in the "physical information" column are included.

According to the table data TDA of fig. 3, for example, in the case where the physical information output from the main apparatus 12 includes an examination result indicating that the helicobacter pylori examination is positive and the physical information does not include an examination result of the pepsinogen examination, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is bacterial infection.

According to the table data TDA of fig. 3, for example, in the case where the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is positive and the physical information includes a test result indicating that the pepsinogen test is negative, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is ulcer.

According to the table data TDA of fig. 3, for example, in the case where the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is positive and the physical information includes a test result indicating that the pepsinogen test is positive, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is atrophic gastritis.

According to the table data TDA of fig. 3, for example, in the case where the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is negative and the physical information includes a test result indicating that the pepsinogen test is positive, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is atrophic gastritis.

According to the table data TDA of fig. 3, for example, in the case where the body information output from the main body apparatus 12 includes information indicating that the sterilization treatment of helicobacter pylori has been completed, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is the sterilized treatment.

According to the table data TDA of fig. 3, for example, when the physical information output from the main apparatus 12 includes an examination result indicating that there is a mutation in the CDH1 gene, it is specified that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is diffuse gastric cancer.

According to the table data TDA of fig. 3, for example, when the body information output from the main apparatus 12 includes information indicating that the ablation treatment has been performed and an examination result indicating that there is a variation in the HER2 gene, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination has been performed.

Based on the table data TDA in fig. 3, for example, when the body information output from the main apparatus 12 includes an examination result indicating that the EB virus examination is positive, it is specified that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is atrophic gastritis.

According to the table data TDA of fig. 3, for example, when the body information output from the main body apparatus 12 includes information indicating that NSAIDs are being administered, it is determined that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is ulcer.

According to the table data TDA of fig. 3, for example, when the body information output from the main apparatus 12 includes information indicating that there is a self-perceived symptom of the subject, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is an ulcer. In this case, examples of the self-perceived symptoms include pain around the skin, hyperchlorhydria, dark stool, anemia, and the like.

For example, when one abnormal symptom appearing in the stomach of a subject under endoscopic examination cannot be specified due to the reason that the abnormal symptom corresponding to the body information output from the main body apparatus 12 is not included in the table data TDA, the abnormal symptom specifying unit 13A of the present embodiment outputs ambiguous information indicating the presence or absence of the abnormal symptom appearing in the stomach of the subject to the lesion extraction function control unit 13B.

For example, when simultaneously specifying a plurality of abnormal symptoms corresponding to the table data TDA of fig. 3, the abnormal symptom specifying unit 13A of the present embodiment outputs information indicating one of the plurality of abnormal symptoms having the highest priority to the lesion extraction function control unit 13B. The priority levels of the abnormal symptoms included in the table data TDA in fig. 3 may be set to a predetermined order according to the magnitude of risk values, for example, which will be described later.

The lesion extraction function control unit 13B performs processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, and performs control for performing lesion extraction processing in the one lesion extraction unit, based on the information output from the abnormal symptom determination unit 13A.

Specifically, for example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is bacterial infection, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESA specialized for lesion extraction in helicobacter pylori infection from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESA.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is an ulcer, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESB that is specialized in lesion extraction at the time of ulcer discovery from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESB.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is atrophic gastritis, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESC, which is specialized for lesion extraction in atrophic gastritis discovery, from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom has been sterilized, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESD, which is specialized for the lesion extraction after the sterilization of helicobacter pylori, from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESD.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is a diffuse gastric cancer, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESE that is specialized for lesion extraction in the diffuse gastric cancer discovery from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESE.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom has been removed, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESF that is specialized for lesion extraction after gastrectomy from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESF.

For example, when the information output from the abnormal symptom specifying unit 13A includes ambiguous information indicating the presence or absence of an abnormal symptom (when the abnormal symptom specifying unit 33A does not specify an abnormal symptom), the lesion extracting function control unit 13B performs control for selecting a lesion extracting unit LESY that is not specialized for a lesion extracting process for a specific abnormal symptom that may appear in the stomach and performing a lesion extracting process in the lesion extracting unit LESY, instead of selecting one lesion extracting unit LESX from the plurality of lesion extracting units LES included in the lesion extracting function unit 13C.

That is, in the specific example described above, 1 arbitrary one of 6 lesion extraction portions LESA, LESB, LESC, LESD, LESE, and LESF corresponds to the lesion extraction portion LESX.

The lesion extraction function unit 13C performs lesion extraction processing for extracting a lesion candidate region LA from the endoscopic image EG output from the main apparatus 12 in either the lesion extraction unit LESX or the lesion extraction unit LESY under the control of the lesion extraction function control unit 13B, and outputs information capable of specifying the position of the lesion candidate region LA extracted by the lesion extraction processing to the display control unit 13D.

The display control unit 13D performs the following processing based on the information output from the lesion extraction function unit 13C: the position of the lesion candidate area LA in the endoscopic image EG output from the main apparatus 12 is specified, and a marker MG corresponding to a rectangular frame surrounding the specified lesion candidate area LA is generated. Further, the display control unit 13D performs the following processing: the marker MG is added to the endoscopic image EG to generate a display image, and the generated display image is output to the display device 14. That is, the display control unit 13D of the present embodiment performs processing for displaying, on the display device 14, the endoscopic image EG obtained by imaging the diagnostic target organ of the subject under endoscopic examination and the information indicating the position of the lesion candidate region LA extracted by the lesion extraction unit LESX or the lesion extraction unit LESY together. Then, according to the processing of the display control unit 13D, for example, the display image DGA shown in fig. 4 is displayed on the display device 14. Fig. 4 is a diagram showing an example of a display image displayed by the processing of the diagnosis support apparatus according to embodiment 1.

However, for example, in a process of extracting a lesion candidate region from an endoscopic image obtained by imaging the inside of the stomach, the accuracy of extraction of the lesion candidate region tends to vary depending on the state of the gastric mucosa of the subject under endoscopic examination. Specifically, for example, in the gastric mucosa after the sterilization treatment of helicobacter pylori, the extraction accuracy of the process of extracting the lesion candidate region from the endoscopic image tends to be lower than that before the sterilization treatment because the size of the lesion is smaller than that before the sterilization treatment and the color tone of the lesion is not obvious.

In contrast, according to the present embodiment, body information including 1 or more pieces of information that can estimate the state of the stomach (organ to be diagnosed) of the subject under endoscopic examination is input to the abnormal symptom specifying unit 13A, one abnormal symptom based on the body information is specified by the abnormal symptom specifying unit 13A, one lesion extracting unit that is the most suitable for the one abnormal symptom is selected from the plurality of lesion extracting units by the lesion extracting function control unit 13B, and the processing of extracting a lesion candidate region from the endoscopic image by the one lesion extracting unit is performed by the lesion extracting function unit 13C. Therefore, according to the present embodiment, the accuracy of extracting a lesion in a target organ to be diagnosed can be ensured.

Further, according to the present embodiment, for example, the lesion extraction function control unit 13B may be configured to perform processing for setting the sensitivity in accordance with an instruction (instruction of the user) performed by the input I/F123, and perform control for performing lesion extraction processing based on the sensitivity in the lesion extraction unit LESX or the lesion extraction unit LESY. Further, according to the configuration of the lesion extraction function control unit 13B, for example, when the sensitivity of the lesion extraction process performed by the lesion extraction unit LESX or the lesion extraction unit LESY is set to be relatively high, the extraction frequency of the lesion candidate region extracted from the endoscopic image EG can be relatively increased. Further, according to the configuration of the lesion extraction function control unit 13B as described above, for example, when the sensitivity of the lesion extraction process performed by the lesion extraction unit LESX or the lesion extraction unit LESY is set to be relatively low, the probability of the lesion candidate region extracted from the endoscopic image EG (the probability of a lesion actually being included in the lesion candidate region) can be relatively increased. In the present embodiment, the size of the sensitivity of the lesion extraction process is set to a value inversely proportional to the size of a threshold applied to a parameter such as a pixel value acquired from the endoscopic image EG when detecting a lesion candidate region included in the endoscopic image EG.

Further, according to the present embodiment, the lesion extraction process in the lesion extraction unit LESX or the lesion extraction unit LESY is not limited to the process performed on the endoscopic image obtained in the endoscopic examination (output from the main body apparatus 12), and may be a process performed on an endoscopic image obtained from an endoscopic image recorded in advance in a video recording apparatus or the like (not shown).

Further, by appropriately changing the configuration of the present embodiment, for example, it is possible to cope with a case where the organ to be diagnosed of the subject under endoscopic examination is an organ different from the stomach.

In addition, according to the present embodiment, the physical information of the subject who receives the endoscopic examination may include, for example, information indicating at least one of the eating habits of the subject, the smoking history of the subject, the age of the subject, the sex of the subject, the race of the subject, and the family medical history of cancer in the family of the subject.

In addition, according to the present embodiment, instead of the table data TDA shown in fig. 3, for example, when the table data TDB shown in fig. 5 is stored in the storage medium 132, the respective units of the diagnosis assistance apparatus 13 may perform operations corresponding to the table data TDB. The operation and the like of the modified example of the present embodiment will be described below. Hereinafter, for the sake of simplicity, detailed description of portions to which the above-described operations and the like can be applied will be appropriately omitted. Fig. 5 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to the modification example of embodiment 1.

The abnormal symptom identification unit 13A performs processing for identifying one abnormal symptom appearing in the stomach of the subject under endoscopic examination, the size of the lesion risk in the one abnormal symptom, and the recommendation of whether to observe special light for the one abnormal symptom, respectively, based on the body information output from the main body apparatus 12. The abnormal symptom determination unit 13A outputs, to the lesion extraction function control unit 13B, information relating to one abnormal symptom determined by the above-described processing, the size of the lesion risk determined by the above-described processing, and the recommendation of whether or not to observe the one abnormal symptom determined by the above-described processing with special light.

Specifically, the abnormal symptom identification unit 13A reads the table data TDB of fig. 5 from the storage medium 132, for example, and compares the read table data TDB with the body information output from the main body apparatus 12, thereby identifying one abnormal symptom appearing in the stomach of the subject under the endoscopic examination, a risk value preset based on the one abnormal symptom, and a recommendation of whether or not to observe the one abnormal symptom with special light.

The "body information" column of the table data TDB in fig. 5 includes a plurality of items corresponding to body information (information input in the input I/F123) output from the main apparatus 12. In the "abnormal symptom" column of the table data TDB in fig. 5, a plurality of items indicating the symptoms of the stomach corresponding to the respective items in the "physical information" column are included. In addition, the "lesion risk" column of the table data TDB of fig. 5 includes a plurality of items indicating a risk value corresponding to a value set according to the height of the lesion occurrence frequency for each symptom of the stomach in the "abnormal symptom" column. In the column "special light observation" in the table data TDB of fig. 5, a plurality of items indicating whether special light observation (narrow-band light observation) is recommended for each symptom of the stomach in the column "abnormal symptom". In addition, the values of the items in the "lesion risk" column of the table data TDB of fig. 5 are set to relative values greater than 0 and less than 100.

According to the table data TDB of fig. 5, for example, when the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is positive and the physical information does not include a test result of the pepsinogen test, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic test is bacterial infection, the risk value corresponding to the abnormal symptom is 10, and switching to the special light observation is not recommended for the abnormal symptom.

For example, when the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is positive and the physical information includes a test result indicating that the pepsinogen test is negative, the table data TDB of fig. 5 identifies that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is an ulcer, the risk value corresponding to the abnormal symptom is 5, and the switching to the special light observation is not recommended for the abnormal symptom.

For example, when the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is positive and the physical information includes a test result indicating that the pepsinogen test is positive, the table data TDB of fig. 5 identifies that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is atrophic gastritis, the risk value corresponding to the abnormal symptom is 20, and switching to special light observation to the abnormal symptom is not recommended.

For example, when the physical information output from the main apparatus 12 includes a test result indicating that the helicobacter pylori test is negative and the physical information includes a test result indicating that the pepsinogen test is positive, the table data TDB of fig. 5 identifies that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is atrophic gastritis, the risk value corresponding to the abnormal symptom is 30, and switching to special light observation is not recommended for the abnormal symptom.

According to the table data TDB of fig. 5, for example, when the body information output from the main body apparatus 12 includes information indicating that the sterilization treatment of helicobacter pylori has been completed, it is determined that the abnormal symptom occurring in the stomach of the subject under the endoscopic examination is the sterilization treatment, the risk value corresponding to the abnormal symptom is 50, and switching to the special light observation is recommended for the abnormal symptom, respectively.

According to the table data TDB of fig. 5, for example, when the body information output from the main apparatus 12 includes an examination result indicating that there is a mutation in the CDH1 gene, it is determined that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is diffuse gastric cancer, the risk value corresponding to the abnormal symptom is 50, and switching to special light observation is recommended for the abnormal symptom.

According to the table data TDB of fig. 5, for example, when the body information output from the main apparatus 12 includes information indicating that the treatment has been removed and an examination result indicating that there is a mutation in the HER2 gene, it is determined that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is the treatment having been removed, the risk value corresponding to the abnormal symptom is 50, and switching to special light observation is not recommended for the abnormal symptom.

According to the table data TDB of fig. 3, for example, when the body information output from the main apparatus 12 includes an examination result indicating that the EB virus examination is positive, it is determined that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is atrophic gastritis, the risk value corresponding to the abnormal symptom is 10, and switching to the special light observation is not recommended for the abnormal symptom.

According to the table data TDB of fig. 5, for example, when the body information output from the main body apparatus 12 includes information indicating that NSAIDs are being administered, it is determined that an abnormal symptom appearing in the stomach of the subject under the endoscopic examination is an ulcer, the risk value corresponding to the abnormal symptom is 5, and switching to special light observation is not recommended for the abnormal symptom.

According to the table data TDB of fig. 5, for example, when the body information output from the main apparatus 12 includes information indicating that there is a self-perceived symptom of the subject, it is determined that the abnormal symptom appearing in the stomach of the subject under the endoscopic examination is an ulcer, the risk value corresponding to the abnormal symptom is 5, and switching to the special light observation is not recommended for the abnormal symptom.

For example, if one abnormal symptom appearing in the stomach of the subject under endoscopic examination cannot be specified because the abnormal symptom corresponding to the body information output from the main body device 12 is not included in the table data TDB, the abnormal symptom specifying unit 13A of the present modification outputs ambiguous information indicating the presence or absence of the abnormal symptom appearing in the stomach of the subject to lesion extraction function control unit 13B.

In addition, when one abnormal symptom appearing in the stomach of the subject under endoscopic examination cannot be specified from the table data TDB, the abnormal symptom specifying unit 13A of the present modification outputs, to the lesion extraction function control unit 13B, information indicating whether or not special light observation is recommended and a risk value set based on information included in the body information output from the main body device 12. Specifically, for example, when the body information output from the main apparatus 12 includes information indicating that the X-ray image of the stomach shows the presence of the abnormal symptom specifying unit 13A according to the present modification, the abnormal symptom specifying unit 13A outputs information indicating that the risk value is 80 and switching to the special light observation is not recommended to the lesion extraction function control unit 13B.

In addition, the abnormal symptom specifying unit 13A according to the present modification may increase or decrease the risk value acquired from the table data TDB based on the body information output from the main apparatus 12, for example, when the body information includes information indicating whether or not the subject desires to actively detect a lesion or the like.

Specifically, for example, when the body information output from the main apparatus 12 includes information indicating that the subject desires to actively detect a lesion, the abnormal symptom specifying unit 13A according to the present modification may increase the risk value acquired from the table data TDB based on the information. In addition, the abnormal symptom specifying unit 13A according to the present modification may be configured to reduce the risk value acquired from the table data TDB based on the body information output from the main apparatus 12, for example, when the body information includes information indicating that the subject does not desire to actively detect a lesion.

For example, when a plurality of abnormal symptoms corresponding to the table data TDB of fig. 5 are simultaneously identified, the abnormal symptom identification unit 13A of the present modification outputs, to the lesion extraction function control unit 13B, information that associates one abnormal symptom having the highest priority among the plurality of abnormal symptoms, a risk value corresponding to the one abnormal symptom, and a recommendation of whether to observe special light for the one abnormal symptom. The priority levels of the abnormal symptoms included in the table data TDB in fig. 5 may be set to a predetermined order corresponding to the magnitude of the risk value included in the "lesion risk" column, for example.

The lesion extraction function control unit 13B of the present modification performs, based on the information output from the abnormal symptom specifying unit 13A, processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from among a plurality of lesion extraction units included in the lesion extraction function unit 13C, processing for setting a sensitivity corresponding to a risk value (the size of a lesion risk) included in the information, and control for performing lesion extraction processing based on the sensitivity in the one lesion extraction unit. The lesion extraction function control unit 13B of the present modification controls the lesion extraction function unit 13C to output the processing result of the lesion extraction processing in one lesion extraction unit selected as described above and information indicating whether or not special light observation is recommended, which is included in the information output from the abnormal symptom determination unit 13A, in association with each other.

Specifically, for example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is a bacterial infection and information indicating that the risk value is 10, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESA from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs a process of setting the sensitivity corresponding to the risk value to SP, and performs a control of performing a lesion extraction process based on the sensitivity SP in the lesion extraction unit LESA.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is an ulcer and information indicating that the risk value is 5, the lesion extraction function control unit 13B performs processing for selecting the lesion extraction unit LESB from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SP, and performs control for performing lesion extraction processing based on the sensitivity SP in the lesion extraction unit LESB.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is atrophic gastritis and information indicating that the risk value is any of 20 and 30, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESC from among the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs a process of setting the sensitivity corresponding to the risk value to SQ (> SP), and performs a control of performing a lesion extraction process based on the sensitivity SQ in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is atrophic gastritis and information indicating that the risk value is 10, the lesion extraction function control unit 13B performs a process of selecting a lesion extraction unit LESC from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs a process of setting the sensitivity corresponding to the risk value to SP, and performs a control of performing a lesion extraction process based on the sensitivity SP in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom has been sterilized and information indicating that the risk value is 50, the lesion extraction function control unit 13B performs processing for selecting a lesion extraction unit LESD from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SR (> SQ), and performs control for performing lesion extraction processing based on the sensitivity SR in the lesion extraction unit LESD.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is diffuse gastric cancer and information indicating that the risk value is 50, the lesion extraction function control unit 13B performs processing for selecting a lesion extraction unit LESE from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SR, and performs control for performing lesion extraction processing based on the sensitivity SR in the lesion extraction unit LESE.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating that the abnormal symptom is ablation treatment and information indicating that the risk value is 50, the lesion extraction function control unit 13B performs processing for selecting the lesion extraction unit LESF from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value as SR, and performs control for performing lesion extraction processing based on the sensitivity SR in the lesion extraction unit LESF.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating the presence or absence of an abnormal symptom and information indicating that the risk value is 80, the lesion extraction function control unit 13B performs processing for selecting the lesion extraction unit LESY included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SR, and performs control for performing lesion extraction processing based on the sensitivity SR in the lesion extraction unit LESY.

For example, when the information output from the abnormal symptom specifying unit 13A includes information indicating the presence or absence of an abnormal symptom and information indicating that the risk value is 10, the lesion extraction function control unit 13B performs processing for selecting the lesion extraction unit LESY included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SP, and performs control for performing lesion extraction processing based on the sensitivity SP in the lesion extraction unit LESY.

The lesion extraction function unit 13C performs lesion extraction processing for extracting a lesion candidate region LA from the endoscopic image EG output from the main apparatus 12 in either the lesion extraction unit LESX or the lesion extraction unit LESY under the control of the lesion extraction function control unit 13B, and outputs information capable of specifying the position of the lesion candidate region LA extracted by the lesion extraction processing and information indicating whether special light observation is recommended or not specified by the abnormal symptom specification unit 13A to the display control unit 13D.

The display control unit 13D specifies the position of the lesion candidate area LA in the endoscopic image EG output from the main apparatus 12 based on the information output from the lesion extraction function unit 13C, and performs processing for generating a marker MG corresponding to a rectangular frame surrounding the periphery of the specified lesion candidate area LA.

When the information output from the lesion extraction function unit 13C includes information indicating that switching to special light observation is not recommended, the display control unit 13D performs processing for generating a display image similar to the display image DGA illustrated in fig. 4 and outputting the display image to the display device 14.

When the information output from the lesion extraction function unit 13C includes information indicating that switching to special light observation is recommended, the display control unit 13D performs processing for generating character information MJ including a character string or the like for prompting switching from the white light observation mode to the special light observation mode. In addition, when the information output from the lesion extraction function unit 13C includes information indicating that switching to special light observation is recommended, the display control unit 13D performs the following processing: the mark MG is added to the endoscopic image EG, and the character information MJ is added to a display area outside the endoscopic image EG, thereby generating a display image, and the generated display image is output to the display device 14. That is, when it is detected that the recommendation to switch to the special light observation for one abnormal symptom specified by the abnormal symptom specifying unit 13A is made, the display control unit 13D of the present modification performs processing for displaying, on the display device 14, the endoscopic image EG obtained by imaging the organ to be diagnosed of the subject who is subjected to the endoscopic examination, the information indicating the position of the lesion candidate region LA extracted by the lesion extracting unit LESX or the lesion extracting unit LESY, and the information prompting the execution of the special light observation together. Then, according to the processing of the display control unit 13D, for example, the display image DGB shown in fig. 6 is displayed on the display device 14. Fig. 6 is a diagram showing an example of a display image displayed by processing in the diagnosis assistance apparatus according to the modification example of embodiment 1.

Further, according to this modification, instead of displaying the character information MJ on the display device 14, for example, a sound corresponding to the character information MJ may be output from a speaker not shown.

Further, according to the present modification, for example, each lesion extraction unit included in the lesion extraction function unit 13C may perform lesion extraction processing including processing for calculating the likelihood LK of the lesion candidate region LA extracted from the endoscopic image EG, and output information indicating the position of the lesion candidate region LA, information indicating the calculation result of the likelihood LK, and information indicating whether or not special light observation is recommended, which is determined by the abnormal symptom determination unit 13A, to the display control unit 13D. In such a configuration, for example, when the information output from the lesion extraction function unit 13C includes information indicating that switching to special light observation is recommended and the likelihood LK indicated by the information output from the lesion extraction function unit 13C falls within a predetermined range, the display control unit 13D may perform an operation for generating and displaying the character information MJ on the display device 14.

In addition, according to the present modification, for example, at least 1 lesion extraction unit out of the plurality of lesion extraction units included in the lesion extraction function unit 13C may be subdivided in accordance with the sensitivity in the lesion extraction process. Specifically, for example, instead of the lesion extraction unit LESA, the lesion extraction function unit 13C may be provided with a lesion extraction unit LESAP that performs lesion extraction processing specialized for the lesion extraction in the case of helicobacter pylori infection at the sensitivity SP, a lesion extraction unit LESAQ that performs the lesion extraction processing at the sensitivity SQ, and a lesion extraction unit LESAR that performs the lesion extraction processing at the sensitivity SR.

As described above, according to this modification, it is possible to select one lesion extraction unit from among a plurality of lesion extraction units included in the lesion extraction function unit 13C, using a method substantially similar to the method of the above-described embodiment (the method using the table data TDA). Further, according to the present modification, the lesion extraction processing in one lesion extraction unit selected as described above can be performed with sensitivity set according to the frequency of occurrence of lesions in the organ to be diagnosed of the subject under endoscopic examination. Therefore, according to the present modification, the accuracy of extracting a lesion in a target organ to be diagnosed can be ensured.

In addition, according to the present modification, for example, when the body information output from the main body apparatus 12 includes information indicating that the organ to be diagnosed of the subject under the endoscopic examination is the esophagus and the race of the subject is euromate, the abnormal symptom specifying unit 13A may perform a process for specifying whether or not Barrett's esophagus (Barrett's esophagus) is present in the organ to be diagnosed. Further, according to the present modification, for example, when it is determined that one abnormal symptom appearing in the esophagus of the subject under the endoscopic examination is barrett's esophagus, the lesion extracting function control unit 13B may perform a process of selecting one lesion extracting unit corresponding to the one abnormal symptom; performing a process for setting a sensitivity corresponding to the one abnormal symptom; and performing control for performing lesion extraction processing based on the sensitivity in the one lesion extraction unit.

(embodiment 2)

Fig. 7 and 8 are views of embodiment 2.

In this embodiment, detailed description of portions having the same configuration and the like as those of embodiment 1 is appropriately omitted, and portions having configurations and the like different from those of embodiment 1 are mainly described.

The endoscope system 1 of the present embodiment is configured to include a diagnosis assisting apparatus 23 shown in fig. 7, for example, instead of the diagnosis assisting apparatus 13. Fig. 7 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 2.

The diagnosis assisting apparatus 23 has substantially the same configuration as that in which the abnormal symptom specifying unit 13A of the diagnosis assisting apparatus 13 is replaced with the abnormal symptom specifying unit 23A, and the lesion extraction function control unit 13B of the diagnosis assisting apparatus 13 is replaced with the lesion extraction function control unit 23B.

The abnormal symptom identification unit 23A is configured to, when detecting an abnormal finding detection instruction output from the main body apparatus 12, perform processing for identifying one abnormal symptom appearing in the organ to be diagnosed of the subject under endoscopic examination, based on the endoscopic image input at the timing when the abnormal finding detection instruction is detected. The abnormal symptom specifying unit 23A is configured to output information indicating one abnormal symptom specified as described above to the lesion extraction function control unit 23B. That is, the abnormal symptom specifying unit 23A is configured to perform processing for specifying one abnormal symptom appearing in a diagnosis target organ of a subject under endoscopic examination, based on an endoscopic image obtained by imaging the diagnosis target organ.

Specifically, the abnormal symptom specifying unit 23A performs processing for specifying one abnormal symptom using, for example, a feature quantity (at least 1 of hue, chroma, and lightness) indicating a color tone of the endoscopic image output from the main apparatus 12, and a statistic quantity (at least 1 of variance of pixel values and distribution of luminance gradient) calculated based on pixel values of a plurality of pixels included in a region of interest set in the endoscopic image.

Alternatively, for example, when the abnormal symptom specifying unit 23A includes a classifier configured to recognize a plurality of abnormal symptoms that may occur in the predetermined organ by machine learning using an image obtained by imaging the inside of the predetermined organ and classify each abnormal symptom, the abnormal symptom specifying unit performs processing for specifying one abnormal symptom from an output result obtained by inputting a pixel value or the like of the endoscopic image output from the main apparatus 12 to the classifier.

The abnormal symptom specifying unit 23A is configured to continuously output information indicating one abnormal symptom specified based on one abnormal finding detection instruction to the lesion extraction function control unit 23B during a period from a timing when the one abnormal finding detection instruction output from the main body apparatus 12 is detected to a timing when the next abnormal finding detection instruction output from the main body apparatus 12 is detected.

The lesion extraction function control unit 23B is configured to perform processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, and to perform control for performing lesion extraction processing in the one lesion extraction unit, based on the information output from the abnormal symptom specifying unit 23A.

Next, the operation of the present embodiment will be described.

After connecting the respective parts of the endoscope system 1 and turning on the power supply, the user inserts the insertion portion of the endoscope 11 into the subject, and places the distal end portion of the insertion portion at a position where a desired subject in the stomach of the subject can be imaged. The user performs an abnormality finding detection instruction by operating an abnormality finding detection instruction switch of the input I/F123 at a time TP, which is a period during which a desired object inside the stomach of the subject is observed, for example.

When the power of the main body device 12 is turned on, the control unit 124 controls the light source unit 121 to generate B light, G light, and R light as illumination light sequentially or simultaneously. Then, under the control of the control unit 124, illumination light is supplied from the light source unit 121 to the endoscope 11, return light from the subject illuminated with the illumination light is captured by the imaging unit 111, an endoscopic image EG corresponding to the imaging signal output from the imaging unit 111 to the main body apparatus 12 is generated by the image generation unit 122, and the generated endoscopic image EG is sequentially output to the diagnosis assistance apparatus 23 frame by frame.

When detecting that an abnormality detection instruction is given to the input I/F123, the control unit 124 performs an operation for outputting the abnormality detection instruction to the diagnosis assisting apparatus 23.

When detecting an abnormality finding detection instruction output from the main body apparatus 12 at the time point TP, the abnormal symptom identification unit 23A performs processing for identifying one abnormal symptom appearing in the organ to be diagnosed of the subject under endoscopic examination, based on the endoscopic image EG input from the main body apparatus 12 at the time point TP. The abnormal symptom specifying unit 23A outputs information indicating one abnormal symptom specified by the above-described processing to the lesion extraction function control unit 23B.

Specifically, for example, when detecting that the color tone of the endoscopic image EG output from the main apparatus 12 is a high-chroma red color tone (corresponding to the color tone of a red mucous membrane), the abnormal symptom specifying unit 23A specifies that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is bacterial infection.

For example, when detecting that the color tone of the endoscopic image EG outputted from the main apparatus 12 is a low-chroma red color tone (corresponding to a discolored color tone of mucous membrane), the abnormal symptom specifying unit 23A specifies that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is bacterial infection.

For example, when detecting that the color tone of the endoscopic image EG output from the main apparatus 12 is white (color tone due to white turbid mucus), the abnormal symptom specifying unit 23A specifies that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is bacterial infection.

Further, for example, when the endoscopic image EG output from the main apparatus 12 is subjected to a process using the method disclosed in japanese patent application laid-open No. 2015-181594 and the like, and the presence of black or white fur in the endoscopic image EG is detected, the abnormal symptom determination unit 23A determines that the abnormal symptom occurring in the stomach of the subject under endoscopic examination is ulcer.

For example, when the endoscopic image EG output from the main apparatus 12 is subjected to a predetermined process and the presence of a dendritic blood vessel in the endoscopic image EG is detected, the abnormal symptom specifying unit 23A specifies that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is atrophic gastritis.

That is, one abnormal symptom appearing in the organ to be diagnosed of the subject under endoscopic examination is specified based on the endoscopic image EGP output from the main apparatus 12 at the time point TP based on the operation of the abnormal symptom specifying unit 23A and the like described above, and information indicating the specified one abnormal symptom is output to the lesion extraction function control unit 23B.

For example, when one abnormal symptom cannot be specified from the endoscopic image EG output from the main apparatus 12, the abnormal symptom specifying unit 23A outputs ambiguous information indicating the presence or absence of the abnormal symptom appearing in the stomach of the subject to the lesion extraction function control unit 23B.

The abnormal symptom specifying unit 23A continuously outputs information indicating one abnormal symptom specified based on one abnormal finding detection instruction to the lesion extraction function control unit 23B during a period from a time TP when the one abnormal finding detection instruction output from the main body apparatus 12 is detected to a time TQ when a next abnormal finding detection instruction output from the main body apparatus 12 is detected.

The lesion extraction function control unit 23B performs processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, and performs control for performing lesion extraction processing in the one lesion extraction unit, based on the information output from the abnormal symptom determination unit 23A.

Specifically, for example, when the information output from the abnormal symptom specifying unit 23A includes information indicating that the abnormal symptom is a bacterial infection, the lesion extraction function control unit 23B performs a process of selecting a lesion extraction unit LESA from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESA.

For example, when the information indicating that the abnormal symptom is ulcer is included in the information output from the abnormal symptom specifying unit 23A, the lesion extraction function control unit 23B performs a process of selecting the lesion extraction unit LESB from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing the lesion extraction process in the lesion extraction unit LESB.

For example, when the information output from the abnormal symptom specifying unit 23A includes information indicating that the abnormal symptom is atrophic gastritis, the lesion extraction function control unit 23B performs a process of selecting a lesion extraction unit LESC from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, and performs a control of performing a lesion extraction process in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 23A includes ambiguous information indicating the presence or absence of an abnormal symptom (when the abnormal symptom specifying unit 23A does not specify an abnormal symptom), the lesion extraction function control unit 23B performs a process of selecting the lesion extraction unit LESY included in the lesion extraction function unit 13C and a control of performing a lesion extraction process in the lesion extraction unit LESY.

That is, in the specific example described above, 1 arbitrary one of the 3 lesion extraction portions LESA, LESB, and LESC corresponds to the lesion extraction portion LESX.

The lesion extraction function unit 13C performs lesion extraction processing for extracting a lesion candidate region LAP from the endoscopic image EGP output from the main body apparatus 12 in either the lesion extraction unit LESX or the lesion extraction unit LESY under the control of the lesion extraction function control unit 23B, and outputs information capable of specifying the position of the lesion candidate region LAP extracted by the lesion extraction processing to the display control unit 13D.

The display control unit 13D specifies the position of the lesion candidate area LAP in the endoscopic image EGP input at the time point TP based on the information output from the lesion extraction function unit 13C, and performs processing for generating a marker MGP corresponding to a frame of a rectangle surrounding the specified lesion candidate area LAP. The display control unit 13D performs processing for generating a display image by adding the marker MGP to the endoscopic image EGP and outputting the generated display image to the display device 14. That is, the display control unit 13D of the present embodiment performs processing for displaying, on the display device 14, the endoscopic image EGP obtained by imaging the diagnostic target organ of the subject under endoscopic examination together with information indicating the position of the lesion candidate region LAP extracted by the lesion extraction unit LESX or the lesion extraction unit LESY. Then, according to the processing of the display control unit 13D, the display image DGC shown in fig. 8 is displayed on the display device 14, for example. Fig. 8 is a diagram showing an example of a display image displayed by the processing of the diagnosis support apparatus according to embodiment 2.

As described above, according to the present embodiment, an endoscopic image obtained by imaging the inside of the stomach (organ to be diagnosed) of a subject to be examined is input to the abnormal symptom specifying unit 23A, one abnormal symptom based on the processing result of the endoscopic image is specified by the abnormal symptom specifying unit 23A, one lesion extraction unit most suitable for the one abnormal symptom is selected from the plurality of lesion extraction units by the lesion extraction function control unit 23B, and the processing of extracting a lesion candidate region from the endoscopic image by the one lesion extraction unit is performed by the lesion extraction function unit 13C. Therefore, according to the present embodiment, the accuracy of extracting a lesion in a target organ to be diagnosed can be ensured.

As described above, according to the present embodiment, the abnormal symptom specifying unit 23A performs processing for specifying one abnormal symptom appearing in the stomach of the subject based on the endoscopic image at each timing when the input I/F123 instructs the abnormality finding detection. Therefore, according to the present embodiment, for example, in a situation where the user observes a desired observation target site in the stomach such as the cardiac portion, the fundus portion, the corpus gastri portion, and the pylorus portion, by operating the abnormality finding detection instruction switch of the input I/F123 at an appropriate timing, it is possible to extract a lesion corresponding to an abnormal symptom appearing in the desired observation target site with high accuracy.

In addition, according to the present embodiment, for example, when the organ to be diagnosed of the subject under endoscopy is the stomach, the abnormal symptom specifying unit 23A may perform processing for specifying one abnormal symptom other than bacterial infection, ulcer, and atrophic gastritis, and the lesion extraction function control unit 23B may perform processing for selecting one lesion extraction unit corresponding to the one abnormal symptom, and may perform control for performing the lesion extraction processing in the one lesion extraction unit.

Further, according to the present embodiment, for example, when the organ to be diagnosed of the subject under endoscopy is the esophagus, the abnormal symptom specifying unit 23A may specify that one abnormal symptom appearing in the organ to be diagnosed is the barrett's esophagus from the endoscopic image EG output from the main apparatus 12, and the lesion extraction function control unit 23B may perform the processing for selecting one lesion extraction unit corresponding to the one abnormal symptom, and perform the control for performing the lesion extraction processing in the one lesion extraction unit.

(embodiment 3)

Fig. 9 and 10 are views of embodiment 3.

In this embodiment, detailed description of a portion having the same configuration and the like as at least one of embodiments 1 and 2 is appropriately omitted, and a portion having a configuration and the like different from those of embodiments 1 and 2 is mainly described.

The endoscope system 1 of the present embodiment is configured to include a diagnosis assisting apparatus 33 shown in fig. 9, for example, instead of the diagnosis assisting apparatus 13. Fig. 9 is a block diagram for explaining a specific example of the configuration of the diagnosis support apparatus according to embodiment 3.

The diagnosis assisting apparatus 33 has substantially the same configuration as that in which the abnormal symptom specifying unit 13A of the diagnosis assisting apparatus 13 is replaced with the abnormal symptom specifying unit 33A and the lesion extraction function control unit 13B of the diagnosis assisting apparatus 13 is replaced with the lesion extraction function control unit 33B.

The abnormal symptom specifying unit 33A is configured to perform processing for specifying one abnormal symptom appearing in a diagnosis target organ of a subject under endoscopic examination, based on the body information output from the main body apparatus 12.

The abnormal symptom specifying unit 33A is configured to perform processing for specifying an observation target site corresponding to an anatomical site of a diagnosis target organ included in the endoscopic image output from the main apparatus 12. In other words, the observation target region is specified as one of a plurality of regions in the case of anatomically classifying the diagnosis target organ.

Specifically, for example, when a classifier is provided, the abnormal symptom specifying unit 33A performs processing for specifying a site to be observed from an output result obtained by inputting a pixel value or the like of an endoscopic image output from the main body apparatus 12 to the classifier, and the abnormal symptom specifying unit recognizes and classifies a plurality of sites inside a predetermined organ for each site by machine learning using an image obtained by imaging the plurality of sites.

Further, according to the present embodiment, for example, when an angle knob (not shown) capable of performing an operation for changing the angle of the distal end portion of the insertion portion of the endoscope 11 is provided in the operation portion of the endoscope 11, the abnormal symptom specifying unit 33A may perform a process for specifying the observation target site using a detection result obtained by detecting the operation state of the angle knob.

The abnormal symptom identification unit 33A is configured to perform, based on the one abnormal symptom identified by the above-described processing and the observation target site, processing for identifying the magnitude of the lesion risk corresponding to the one abnormal symptom appearing in the observation target site and the recommendation of whether or not to perform special light observation for the one abnormal symptom appearing in the observation target site. The abnormal symptom specifying unit 33A is configured to output, to the lesion extraction function control unit 33B, information associating one abnormal symptom specified by the above-described processing, the size of the lesion risk specified by the above-described processing, and the recommendation of whether or not to observe the special light specified by the above-described processing.

The lesion extraction function control unit 33B is configured to perform processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, based on the information output from the abnormal symptom specifying unit 33A, perform processing for setting a sensitivity corresponding to the size of a lesion risk included in the information, and perform control for performing lesion extraction processing based on the sensitivity in the one lesion extraction unit. The lesion extraction function controller 33B is configured to control the lesion extraction function unit 13C to output the processing result of the lesion extraction processing in the one lesion extraction unit selected as described above in association with information indicating whether or not special light observation is recommended, the information being included in the information output from the abnormal symptom determiner 33A.

Next, the operation of the present embodiment will be described.

After connecting the respective parts of the endoscope system 1 and turning on the power supply, the user inserts the insertion portion of the endoscope 11 into the subject, and places the distal end portion of the insertion portion at a position where a desired object in the stomach of the subject can be imaged.

When the power of the main body device 12 is turned on, the control unit 124 controls the light source unit 121 to generate B light, G light, and R light as illumination light sequentially or simultaneously. Then, under the control of the control unit 124, illumination light is supplied from the light source unit 121 to the endoscope 11, return light from the subject illuminated with the illumination light is captured by the imaging unit 111, an endoscopic image EG corresponding to the imaging signal output from the imaging unit 111 to the main body device 12 is generated by the image generation unit 122, and the generated endoscopic image EG is sequentially output to the diagnosis assistance device 33 frame by frame.

The abnormal symptom specifying unit 33A performs a process for specifying one abnormal symptom appearing in the stomach of the subject under the endoscopic examination based on the body information output from the main body apparatus 12.

Specifically, the abnormal symptom specifying unit 33A reads the table data TDA shown in fig. 3 from the storage medium 132, and compares the read table data TDA with the body information output from the main body apparatus 12 to specify one abnormal symptom appearing in the stomach of the subject under the endoscopic examination.

The abnormal symptom specifying unit 33A performs the following processing: an observation target region in the stomach of a subject to be examined under an endoscopic examination is specified from an endoscopic image output from the main apparatus 12.

Specifically, the abnormal symptom specifying unit 33A specifies to which of the stomach, vestibular, fornix, gastric corner, and other parts the observation target part in the stomach belongs, for example, based on the endoscopic image output from the main apparatus 12.

The abnormal symptom specifying unit 33A performs the following processing: based on one abnormal symptom and the observation target site determined by the aforementioned processing, the magnitude of the lesion risk corresponding to the one abnormal symptom appearing at the observation target site and the recommendation of special light observation for the one abnormal symptom appearing at the observation target site are determined, respectively. The abnormal symptom determination unit 33A outputs information associating one abnormal symptom determined by the aforementioned processing, the size of the lesion risk determined by the aforementioned processing, and the recommendation of whether or not to observe the special light determined by the aforementioned processing to the lesion extraction function control unit 33B.

Specifically, the abnormal symptom identification unit 33A reads the table data TDC shown in fig. 10 from the storage medium 132, for example, and identifies the magnitude of the lesion risk corresponding to the one abnormal symptom appearing in the observation target site and the recommendation of the special light observation for the one abnormal symptom appearing in the observation target site by comparing the read table data TDC with the one abnormal symptom identified by the above-described processing and the observation target site. Fig. 10 is a diagram showing an example of table data used for processing by the diagnosis support apparatus according to embodiment 3.

The column of "abnormal symptom" in the table data TDC in fig. 10 includes a plurality of items obtained by subdividing at least 1 of the stomach symptoms included in the column of "abnormal symptom" in the table data TDA in fig. 3 according to the stomach region. In the column "detailed part" of the table data TDC in fig. 10, a plurality of items indicating the parts of the stomach corresponding to the respective items in the column "abnormal symptom" are included. In addition, the "lesion risk" column of the table data TDC in fig. 10 includes a plurality of items indicating a risk value corresponding to a value set according to the magnitude of the occurrence frequency of lesions for each combination of the stomach symptoms in the "abnormal symptom" column and the stomach parts in the "detailed part" column. In addition, the column of "special light observation" in the table data TDC of fig. 10 includes a plurality of items indicating whether special light observation (narrow-band light observation) is required for the stomach symptom in the column of "abnormal symptom" in the stomach portion in the column of "detailed portion". In addition, the values of the items in the "lesion risk" column of the table data TDC of fig. 10 are set to relative values larger than 0 and smaller than 100.

According to the table data TDC of fig. 10, for example, when it is determined that each abnormal symptom appearing in the stomach of the subject under the endoscopic examination is atrophic gastritis and the observation target portion in the stomach is either the stomach portion or the vestibule portion, the risk value corresponding to the abnormal symptom appearing in the observation target portion is 25 and switching to the special light observation is not recommended for the abnormal symptom appearing in the observation target portion.

According to the table data TDC of fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is atrophic gastritis and that an observation target portion in the stomach is a fornix portion, it is determined that a risk value corresponding to the abnormal symptom appearing in the observation target portion is 8 and switching to special light observation is not recommended for the abnormal symptom appearing in the observation target portion.

According to the table data TDC of fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is atrophic gastritis and that an observation target portion in the stomach is another portion not conforming to any of the stomach portion, the vestibule portion, and the fornix portion, it is determined that the risk value corresponding to the abnormal symptom appearing in the observation target portion is 20 and that switching to special light observation is not recommended for the abnormal symptom appearing in the observation target portion.

From the table data TDC in fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is an ulcer and it is determined that an observation target portion in the stomach is either a gastric corner portion or a gastric body portion, it is determined that a risk value corresponding to the abnormal symptom appearing in the observation target portion is 10 and it is not recommended to switch to special light observation for the abnormal symptom appearing in the observation target portion.

From the table data TDC in fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is an ulcer and it is determined that an observation target portion in the stomach is another portion not conforming to either of the corner portion and the body portion of the stomach, it is determined that the risk value corresponding to the abnormal symptom appearing in the observation target portion is 5 and that switching to special light observation is not recommended for the abnormal symptom appearing in the observation target portion.

Based on the table data TDC in fig. 10, for example, when it is specified that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is diffuse gastric cancer and an observation target portion in the stomach is a gastric corner portion, it is specified that a risk value corresponding to the abnormal symptom appearing in the observation target portion is 50 and it is recommended to switch the abnormal symptom appearing in the observation target portion to special light observation.

Based on the table data TDC in fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is diffuse gastric cancer and an observation target portion in the stomach is a gastric body portion, it is determined that a risk value corresponding to the abnormal symptom appearing in the observation target portion is 40 and it is recommended to switch the abnormal symptom appearing in the observation target portion to special light observation.

According to the table data TDC of fig. 10, for example, when it is determined that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is diffuse gastric cancer and an observation target portion in the stomach is a dome portion, it is determined that the risk value corresponding to the abnormal symptom appearing in the observation target portion is 20 and switching to special light observation is not recommended for the abnormal symptom appearing in the observation target portion.

According to the table data TDC of fig. 10, for example, when it is specified that an abnormal symptom appearing in the stomach of a subject under endoscopic examination is a diffuse stomach cancer and it is specified that an observation target portion in the stomach is another portion not conforming to any of the corner portion, the body portion, and the dome portion, it is specified that the risk value corresponding to the abnormal symptom appearing in the observation target portion is 35 and that switching to special light observation is recommended for the abnormal symptom appearing in the observation target portion.

That is, based on the body information output from the main body device 12 and the endoscopic image EGQ output from the main body device 12, the above-described operation of the abnormal symptom specifying unit 33A specifies one abnormal symptom appearing in the organ to be diagnosed of the subject under endoscopic examination, the size of the lesion risk corresponding to the one abnormal symptom appearing in the observation target region in the diagnosis target organ, and the recommendation of special light observation for the one abnormal symptom appearing in the observation target region.

For example, when one abnormal symptom appearing in the stomach of the subject under endoscopic examination cannot be specified because the abnormal symptom corresponding to the body information output from the main body apparatus 12 is not included in the table data TDA or the like, the abnormal symptom specifying unit 33A outputs ambiguous information indicating the presence or absence of the abnormal symptom appearing in the stomach of the subject to the lesion extraction function control unit 33B.

The lesion extraction function control unit 33B performs processing for selecting one lesion extraction unit corresponding to one abnormal symptom included in the information from the plurality of lesion extraction units included in the lesion extraction function unit 13C, processing for setting a sensitivity corresponding to a risk value (the size of lesion risk) included in the information, and control for performing lesion extraction processing based on the sensitivity in the one lesion extraction unit, based on the information output from the abnormal symptom determination unit 33A. The lesion extraction function controller 33B of the present modification controls the lesion extraction function unit 13C to output the processing result of the lesion extraction processing in one lesion extraction unit selected as described above and information indicating whether or not special light observation is recommended, which is included in the information output from the abnormal symptom determiner 33A, in association with each other.

Specifically, for example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is atrophic gastritis and information indicating that the risk value is 20 or 25, the lesion extraction function control unit 33B performs processing for selecting a lesion extraction unit LESC from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SM, and performs control for performing lesion extraction processing based on the sensitivity SM in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is atrophic gastritis and information indicating that the risk value is 8, the lesion extraction function control unit 33B performs processing for selecting a lesion extraction unit LESC from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SK (< SM), and performs control for performing lesion extraction processing based on the sensitivity SK in the lesion extraction unit LESC.

For example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is an ulcer and information indicating that the risk value is 10, the lesion extraction function control unit 33B performs processing for selecting the lesion extraction unit LESB from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SL (where SK < SL < SM is satisfied), and performs control for performing lesion extraction processing based on the sensitivity SL in the lesion extraction unit LESB.

For example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is an ulcer and information indicating that the risk value is 5, the lesion extraction function control unit 33B performs processing for selecting the lesion extraction unit LESB from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SK, and performs control for performing lesion extraction processing based on the sensitivity SK in the lesion extraction unit LESB.

For example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is diffuse gastric cancer and information indicating that the risk value is any one of 35, 40, and 50, the lesion extraction function control unit 33B performs processing for selecting a lesion extraction unit LESE from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SN (> SM), and performs control for performing lesion extraction processing based on the sensitivity SN in the lesion extraction unit LESE.

For example, when the information output from the abnormal symptom specifying unit 33A includes information indicating that the abnormal symptom is diffuse gastric cancer and information indicating that the risk value is 20, the lesion extraction function control unit 33B performs processing for selecting the lesion extraction unit LESE from the plurality of lesion extraction units LES included in the lesion extraction function unit 13C, performs processing for setting the sensitivity corresponding to the risk value to SM, and performs control for performing lesion extraction processing based on the sensitivity SM in the lesion extraction unit LESE.

For example, when the information output from the abnormal symptom specifying unit 33A includes ambiguous information indicating the presence or absence of an abnormal symptom (when the abnormal symptom specifying unit 33A does not specify an abnormal symptom), the lesion extraction function control unit 33B performs processing for selecting the lesion extraction unit LESY included in the lesion extraction function unit 13C, performs processing for setting one sensitivity corresponding to a default value among the sensitivities SK, SL, SM, and SN, and performs control for performing lesion extraction processing based on the one sensitivity in the lesion extraction unit LESE.

That is, in the specific example described above, 1 arbitrary one of 3 lesion extraction portions LESB, LESC, and LESE corresponds to the lesion extraction portion LESX.

The lesion extraction function unit 13C performs lesion extraction processing for extracting a lesion candidate region LA from the endoscopic image EG output from the main apparatus 12 in either the lesion extraction unit LESX or the lesion extraction unit LESY under the control of the lesion extraction function control unit 33B, and outputs information capable of specifying the position of the lesion candidate region LA extracted by the lesion extraction processing to the display control unit 13D.

The display control unit 13D specifies the position of the lesion candidate area LA in the endoscopic image EG output from the main apparatus 12 based on the information output from the lesion extraction function unit 13C, and performs processing for generating a marker MG corresponding to a rectangular frame surrounding the specified lesion candidate area LA. The display control unit 13D generates a display image by adding the marker MG to the endoscopic image EG, and performs processing for outputting the generated display image to the display device 14. Then, according to the processing of the display control unit 13D, the same display image as the display image DGA illustrated in fig. 4 is displayed on the display device 14.

As described above, according to the present embodiment, it is possible to select one lesion extraction unit from a plurality of lesion extraction units included in the lesion extraction function unit 13C by a method substantially similar to the method described in embodiment 1 (the method using the table data TDA). Further, according to the present embodiment, the lesion extraction processing in one lesion extraction unit selected as described above can be performed with a sensitivity set in accordance with the frequency of occurrence of lesions in the observation target region in the diagnosis target organ of the subject under endoscopic examination. Therefore, according to the present embodiment, the accuracy of extracting a lesion in a target organ to be diagnosed can be ensured. Further, according to the present embodiment, for example, it is possible to reliably extract a lesion candidate region in a case where a region having a relatively high occurrence frequency of a predetermined lesion is observed in a diagnostic target organ of a subject under endoscopic examination, and to suppress extraction of a lesion candidate region in a case where a region having a relatively low occurrence frequency of the predetermined lesion is observed.

The abnormal symptom identification unit 33A of the present embodiment may also identify whether the abnormal symptom appearing in the stomach of the subject under endoscopic examination is ulcer or atrophic gastritis by applying the same processing as described in embodiment 2 to the endoscopic image EG output from the main apparatus 12, for example. The abnormal symptom identification unit 33A of the present embodiment may identify that the abnormal symptom appearing in the stomach of the subject under endoscopic examination is a diffuse gastric cancer by applying, for example, the processing disclosed in japanese patent No. 5242381 to the endoscopic image EG output from the main apparatus 12.

The present invention is not limited to the above-described embodiments, and various modifications and applications can be made without departing from the spirit of the invention.