CN112426119B

CN112426119B - Endoscope screening processing method and device

Info

Publication number: CN112426119B
Application number: CN202110101088.6A
Authority: CN
Inventors: 白蓉; 白银
Original assignee: Shanghai Fuci Medical Technology Co ltd
Current assignee: Shanghai Fuci Medical Technology Co ltd
Priority date: 2021-01-26
Filing date: 2021-01-26
Publication date: 2021-04-13
Anticipated expiration: 2041-01-26
Also published as: CN112426119A

Abstract

The invention provides an endoscope screening processing method and device, which are used for screening a first user by obtaining first image information of an endoscope; obtaining an age of the first user; judging whether the first image information meets a first preset condition, and if so, acquiring first image input data according to the first image information; obtaining first age input data according to the age of the first user; inputting the first image input data and the first age input data into a first training model; obtaining first output information of the first training model; obtaining historical diagnostic information of the first user; according to the historical diagnosis information and the first output information, a first result is obtained and is represented as a screening result of the first user endoscope, so that the technical effects of providing a more accurate diagnosis result according to differences among different users and helping doctors to better and accurately evaluate the state of illness of patients are achieved.

Description

Endoscope screening processing method and device

Technical Field

The invention relates to the technical field of medical treatment, in particular to a method and a device for screening and processing an endoscope.

Background

The endoscope is a detection instrument integrating traditional optics, ergonomics, precision machinery, modern electronics, mathematics and software into a whole. One has an image sensor, optical lens, light source illumination, mechanical device, etc. that can enter the stomach orally or through other natural orifices. Since a lesion which cannot be displayed by X-ray can be seen by an endoscope, it is very useful for a doctor. For example, with the aid of an endoscopist, an ulcer or tumor in the stomach can be observed, and an optimal treatment plan can be developed accordingly.

However, the applicant of the present invention finds that the prior art has at least the following technical problems:

the problems that diagnosis and lesion identification cannot be accurately carried out on endoscope screening results of patients according to differences among the patients, assistance is limited, and the risks of missed diagnosis and misdiagnosis are high exist in the prior art.

Disclosure of Invention

The embodiment of the invention provides an endoscope screening processing method and device, solves the technical problems that diagnosis and focus identification cannot be accurately carried out on endoscope screening results of patients according to differences among the patients, assistance is limited, and the missed diagnosis and misdiagnosis risks are high in the prior art, achieves the technical effects of providing more accurate diagnosis results according to the differences among different users, helping doctors to better and accurately evaluate the illness states of the patients, providing more scientific treatment selection and improving the working efficiency of the doctors.

In view of the above-mentioned problems, embodiments of the present application are proposed to provide an endoscopic screening processing method and apparatus.

In a first aspect, the present invention provides an endoscopic screening treatment apparatus, the apparatus comprising: a first obtaining unit configured to obtain first image information of an endoscope for a first user; a second obtaining unit configured to obtain an age of the first user; a first judging unit configured to judge whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, obtain first image input data from the first image information; a third obtaining unit configured to obtain first age input data according to the first user age; a first input unit configured to input the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets includes: the first age input data, the first image input data, and identification information identifying whether the first user is normal; a fourth obtaining unit, configured to obtain first output information of the first training model, where the first output information is identification information that identifies whether the first user is normal; a fifth obtaining unit, configured to obtain historical diagnostic information of the first user; a sixth obtaining unit for obtaining a first result from the historical diagnostic information and the first output information, the first result being represented as a screening result of the first user endoscope.

In a second aspect, the present invention provides an endoscopic screening method, the method comprising: obtaining first image information of an endoscope for a first user; obtaining an age of the first user; judging whether the first image information meets a first preset condition or not, and when the first preset condition is met, obtaining first image input data according to the first image information; obtaining first age input data according to the age of the first user; inputting the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets comprises: the first age input data, the first image input data, and identification information identifying whether the first user is normal; obtaining first output information of the first training model, wherein the first output information is identification information for identifying whether the first user is normal or not; obtaining historical diagnostic information of the first user; obtaining a first result from the historical diagnostic information and the first output information, the first result being represented as a screening result of the first user endoscope.

In a third aspect, the present invention provides an endoscope screening processing apparatus comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executing the program implementing the steps of the apparatus of any one of the preceding first aspects.

One or more technical solutions in the embodiments of the present application have at least one or more of the following technical effects:

the embodiment of the invention provides an endoscope screening processing method and device, wherein the method comprises the following steps: obtaining first image information of an endoscope for a first user; obtaining an age of the first user; judging whether the first image information meets a first preset condition or not, and when the first preset condition is met, obtaining first image input data according to the first image information; obtaining first age input data according to the age of the first user; inputting the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets comprises: the first age input data, the first image input data, and identification information identifying whether the first user is normal; obtaining first output information of the first training model, wherein the first output information is identification information for identifying whether the first user is normal or not; obtaining historical diagnostic information of the first user; according to the historical diagnosis information and the first output information, a first result is obtained, and the first result is represented as the screening result of the endoscope of the first user, so that the technical problems that diagnosis and lesion identification cannot be accurately carried out on the screening result of the endoscope of the patient according to differences among the patients, the assistance effect is limited, and the missed diagnosis and misdiagnosis risks are high in the prior art are solved, the more accurate diagnosis result is provided according to the differences among different users, a doctor is helped to better accurately evaluate the illness state of the patient, more scientific treatment selection is provided, and the technical effect of improving the working efficiency of the doctor is achieved.

The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.

Drawings

FIG. 1 is a schematic flow chart of a method of endoscopic screening in an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of an endoscopic screening device in an embodiment of the present invention;

fig. 3 is a schematic structural diagram of another exemplary electronic device in an embodiment of the present invention.

Description of reference numerals: a first obtaining unit 11, a second obtaining unit 12, a first judging unit 13, a third obtaining unit 14, a first input unit 15, a fourth obtaining unit 16, a fifth obtaining unit 17, a sixth obtaining unit 18, a bus 300, a receiver 301, a processor 302, a transmitter 303, a memory 304, and a bus interface 306.

Detailed Description

The embodiment of the invention provides an endoscope screening processing method and device, which are used for solving the technical problems that diagnosis and focus identification cannot be accurately carried out on endoscope screening results of patients according to differences among the patients, the assistance effect is limited, and the missed diagnosis and misdiagnosis risks are high in the prior art, so that the more accurate diagnosis results are provided according to the differences among different users, doctors are helped to better and accurately evaluate the illness states of the patients, more scientific treatment selection is provided, and the working efficiency of the doctors is improved.

Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be apparent that the described embodiments are merely some embodiments of the present application and not all embodiments of the present application, and it should be understood that the present application is not limited to the example embodiments described herein.

Summary of the application: the endoscope is a detection instrument integrating traditional optics, ergonomics, precision machinery, modern electronics, mathematics and software into a whole. However, endoscopes are invasive, and in the case of a gastroscope, the patient's discomfort is inevitably caused by the taking of the gastroscope from the mouth, to the esophagus, to the stomach, through the pylorus, and to the duodenum, which is unacceptable for many patients. Although this is related to the sensitivity and tolerance of the patient, it is also inseparable from the skill of the physician, and each step of the endoscopic procedure is likely to cause injury to the patient.

In order to solve the technical problems, the technical scheme provided by the invention has the following general idea:

an embodiment of the present application provides an endoscope screening processing apparatus, the apparatus including: a first obtaining unit configured to obtain first image information of an endoscope for a first user; a second obtaining unit configured to obtain an age of the first user; a first judging unit configured to judge whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, obtain first image input data from the first image information; a third obtaining unit configured to obtain first age input data according to the first user age; a first input unit configured to input the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets includes: the first age input data, the first image input data, and identification information identifying whether the first user is normal; a fourth obtaining unit, configured to obtain first output information of the first training model, where the first output information is identification information that identifies whether the first user is normal; a fifth obtaining unit, configured to obtain historical diagnostic information of the first user; a sixth obtaining unit for obtaining a first result from the historical diagnostic information and the first output information, the first result being represented as a screening result of the first user endoscope.

After the fundamental principle of the present application is introduced, the technical solutions of the present invention are described in detail with reference to the accompanying drawings and specific embodiments, and it should be understood that the specific features in the embodiments and examples of the present application are detailed descriptions of the technical solutions of the present application, and are not limitations of the technical solutions of the present application, and the technical features in the embodiments and examples of the present application may be combined with each other without conflict.

Example one

Fig. 1 is a schematic flow chart of an endoscopic screening processing method according to an embodiment of the present invention. As shown in fig. 1, an embodiment of the present invention provides an endoscopic screening processing method, including:

step 100: obtaining first image information of an endoscope for a first user;

step 200: obtaining an age of the first user;

specifically, the first user is a patient who is subjected to an endoscopic examination in a hospital, the first image information is an image result obtained by endoscopic screening of the user, and the examination result of the first user can be obtained by analyzing the first image. For example, examination of gastrointestinal tract diseases, examination of pancreas, biliary tract diseases, laparoscopy, examination of respiratory tract diseases, examination of urinary tract, and the like can be performed on a patient endoscopically. Furthermore, the age of the first user can be obtained, so that the first image and the age information can be combined to perform more accurate diagnosis and analysis on the illness state of the user in a later period.

Step 300: judging whether the first image information meets a first preset condition or not, and when the first preset condition is met, obtaining first image input data according to the first image information;

further, it is determined whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, first image input data is obtained according to the first image information, step 300 of this embodiment of the present application further includes:

step 310: obtaining first length information of the endoscope penetrating into the first user;

step 320: obtaining height information of the first user;

step 330: obtaining a first arrival result according to the first length information and the height information;

step 340: and when the first arrival result meets a second preset condition, obtaining first image input data according to the first image information.

Specifically, after the first image is obtained, it is then determined whether the first image meets a first predetermined condition, and if the quality of the first image meets a predetermined quality requirement, that is, whether the definition, the gray scale, the size, the pixels, and the like of the first image are in the predetermined requirement, and when the first predetermined condition is determined to be met, the first image input data can be obtained according to the first image, where the specific determination manner is: firstly, obtaining first length information of an endoscope penetrating into a first user, wherein the first length information is the distance of an endoscope device entering a user body in the actual examination process of the endoscope, and the lengths of different items required to extend into the user body are different, such as different positions and different depths of intestinal tract examination and gastroscopy extending into the human body; then, height information of a first user can be obtained, then the first length information and the height information are comprehensively analyzed, and a first arrival result is obtained, namely, for users with different heights, when the same endoscopy is carried out, the depths of the users which need to stretch into the human body are different, compared with a high sub-user and a low sub-user, the stretching length of the sub-user with the higher sub-user stretching length is higher than that of the sub-user with the lower sub-user stretching length, therefore, after the first arrival result is obtained, whether the first arrival result meets a second preset condition is judged, namely whether the arrival position of the user is accurate or not, whether the arrival position has a deviation with a preset position or not is judged according to the height of the first user, and the like; and when the first arrival result meets a second preset condition, the first image input data can be obtained according to the first image, so that the accuracy of model training is improved subsequently, the problem that whether the arrival position is correct or not is judged by a doctor by naked eyes or by experience in the prior art is solved, the purpose of intelligently judging whether the arrival position is correct or not according to the height and the extension length is achieved, the diagnosis accuracy is further improved, and reliable evidence is provided for the doctor to treat patients.

Further, the determining whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, obtaining first image input data according to the first image information, step 300 in this embodiment of the present application further includes:

step 350: acquiring target image information of a detection position according to the subject information;

step 360: inputting the first image information and the target image information into a second training model, wherein the second training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets includes: the first image information, the target image information and an identification result for identifying whether the first image information meets the consistency of the target image information;

step 380: obtaining second output information of the second training model, wherein the second output information identifies whether the first image information meets an identification result of target image information consistency;

step 390: and when the second output information is yes, obtaining first image input data according to the first image.

Specifically, after the inspection subject information is obtained, target image information of the detection position can be obtained, the target image information is standard image information for the detection position, the standard image information is the existing image information related to the detection position, the standard image information can be acquired through big data and used for comparison in practical application, then after the first image information of the endoscope is obtained, the first image information and the second image information are processed through a second training model, and whether the first image information meets the identification result information of the consistency of the target image information or not can be obtained. That is to say, the first image information of the user is compared with the target image information, and whether the first image information meets the consistency of the target image information is further judged.

Furthermore, the training model is a neural network model in the machine learning model, and the machine learning model can continuously learn through a large amount of data, further continuously correct the model, and finally obtain satisfactory experience to process other data. The machine model is obtained by training a plurality of groups of training data, and the process of training the neural network model by the training data is essentially the process of supervised learning. The training model in the embodiment of the application is obtained by utilizing machine learning training through a plurality of groups of training data, and each group of training data in the plurality of groups comprises: the image processing apparatus includes first image information, target image information, and an identification result identifying whether the first image information satisfies consistency of the target image information.

Further, an identification result of whether the first image information meets the consistency of the target image information is used as supervision data, the supervision data is input into each group of training data, supervision learning is carried out on the first image information and the target image information, the identification result information of whether the first image information meets the consistency of the target image information is compared with an output result of a training model, and when the first image information meets the consistency of the target image information, the supervision learning of the group of data is finished, and the supervision learning of the next group of data is carried out; when the image information is inconsistent with the target image information, the training model carries out self-correction until the output result is consistent with the identification result information indicating whether the first image information of the identification meets the target image information consistency, the group of supervised learning is finished, and the next group of data supervised learning is carried out; and (4) through supervised learning of a large amount of data, enabling the output result of the machine learning model to reach a convergence state, and finishing the supervised learning. Through the process of supervising and learning the training model, whether the first image information output by the training model meets the identification result information of the consistency of the target image information is more accurate, the identification result of whether the first image information meets the consistency of the target image information is accurately obtained, and the phenomenon that the detection result of a user is influenced due to inaccurate images is avoided.

Step 400: obtaining first age input data according to the age of the first user;

further, obtaining first age input data according to the age of the first user, in step 400 of this embodiment of the present application, further includes:

step 410: obtaining subject information for an endoscopic examination;

step 420: acquiring first age group information according to the subject information, wherein the first age group information is probability originating age group information of the subject information;

step 430: and subtracting the first age group information from the first user age to obtain first age input data.

Specifically, after the age of the first user is obtained, it is then necessary to analyze the age of the first user and obtain first age input data. Specifically, the method comprises the following steps: first, subject information of an endoscopic examination, that is, examination items performed by a user, such as stomach, duodenum, large intestine, bladder, etc., is obtained, and first age group information is obtained from the subject information based on big data, where the first age group information is probability originating age group information corresponding to the examination subjects of the endoscope, that is, the probability of good onset of disease is different for different examination subjects, for example, a young person after 30 years old is irregular due to work and rest diet, good onset of stomach illness is less before 30 years old, liver disease is good onset for middle aged and elderly people after 50 years old, etc., the probability of obtaining liver disease is less before 50 years old, and first age input data, such as the first user's age being 65 years old, is obtained by subtracting the first age group information from the first user's age, if the good development age is 50 years old, data of 15 years old obtained after subtracting 50 years old from 65 years old can be used as first age input data and further input into the training model, so that the difficulty of machine learning is reduced, the calculation amount is reduced, and the accuracy of a model output result is improved.

Step 500: inputting the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets comprises: the first age input data, the first image input data, and identification information identifying whether the first user is normal;

step 600: obtaining first output information of the first training model, wherein the first output information is identification information for identifying whether the first user is normal or not;

specifically, after the first image input data and the first age input data are obtained, the first image input data and the first age input data may be input into the first training model, and the result of whether the first user is normal or not may be obtained from the output information of the first training model.

Further, as mentioned above, the first training model in the embodiment of the present application is also obtained by training a plurality of sets of training data by machine learning, where each set of training data in the plurality of sets includes: the image processing apparatus includes first age input data, first image input data, and identification information identifying whether a first user is normal.

Further, the identification information of whether the first user is normal is used as supervision data, the supervision data is input into each group of training data, supervision learning is carried out on the first age input data and the first image input data, the identification information of whether the first user is normal is compared with the output result of the training model, when the identification information of whether the first user is normal is consistent with the output result of the training model, the supervision learning of the group of data is finished, and the supervision learning of the next group of data is carried out; when the first user identification information is inconsistent with the first user identification information, the training model carries out self-correction until the output result is consistent with the identification information of whether the first user identification information is normal or not, the group of supervised learning is finished, and the next group of data supervised learning is carried out; and (4) through supervised learning of a large amount of data, enabling the output result of the machine learning model to reach a convergence state, and finishing the supervised learning. Through the process of supervising and learning the training model, the identification information of whether the first user is normal or not output by the training model is more accurate, the accurate identification information of whether the first user is normal or not is obtained, and the influence on the disease diagnosis and the body health of the user caused by the inaccurate judgment of the result of whether the first user is normal or not is avoided.

Step 700: obtaining historical diagnostic information of the first user;

step 800: obtaining a first result from the historical diagnostic information and the first output information, the first result being represented as a screening result of the first user endoscope.

Specifically, the historical diagnosis information of the first user is related disease diagnosis information in the past medical history of the first user, for example, related report results such as clinic diagnosis condition, hospitalization condition, physical examination condition and the like of the user within three years can be collected, so that after the first output information is obtained, the screening result of the endoscope of the first user can be obtained by combining the historical diagnosis information of the user, the accuracy of the screening result of the endoscope of the user is improved, the technical problems that diagnosis and lesion identification cannot be accurately carried out on the screening result of the endoscope of the patient according to differences among the patients in the prior art, the auxiliary effect is limited, the risk of missed diagnosis and misdiagnosis is high are solved, the purposes of intelligently providing more accurate diagnosis results according to the differences among different users and helping doctors to better accurately evaluate the patient are achieved, provides more scientific treatment selection and improves the technical effect of the working efficiency of doctors.

step 3100: obtaining second length information of the endoscope penetrating into the first user;

step 3110: obtaining video stream information of the endoscope in depth to the first user;

step 3120: obtaining N second images according to the video stream information and the second length information;

step 3130: determining whether the N second images are indispensable images of the target image information that reaches the detection position;

step 3140: and when the judgment result is yes, obtaining the first image input data according to the first image information.

Specifically, second length information of the endoscope penetrating into the first user is obtained, wherein the second length information is corresponding length information when the endoscope further reaches a certain position in the body of the user, after an examination subject and a detection position of the user are obtained, target path information of the endoscope extending into the body during detection of the user can be determined accordingly, in the process of image acquisition according to the target path, the endoscope enters the body from the oral cavity and the esophagus, video stream information of the endoscope penetrating into the first user is acquired, then N second images are obtained from the video stream information and the second length information, whether the N second images are essential images of the target image information reaching the detection position is judged, if the essential images are obtained, the path is correct, first image input data can be obtained continuously according to the first image information, and obtaining the first image input data according to the first image information. For example, when the detected position is a stomach, the user needs to enter the esophagus from the pharynx and then reach the stomach during detection, when the vocal cords, vocal folds and the esophagus are marked points, the image information of the marked points can be obtained before reaching the stomach, then whether the image information is the image information of the position which is necessary to reach the stomach is judged, and if the image information is the image information of the position which is necessary to reach the stomach, the subsequent steps can be continued. Therefore, whether the detection position is correct or not is accurately judged, unnecessary body injury to a user caused by position deviation is prevented, and a doctor is helped to improve diagnosis quality and work efficiency.

Further, in order to obtain an accurate endoscope screening result and improve the diagnosis accuracy, step 800 in this embodiment of the present application further includes:

step 810: obtaining first explicit characteristic information according to the subject information, wherein the first explicit characteristic information is human body appearance image information formed by the subject information in historical experience;

step 820: obtaining first appearance image information;

step 830: inputting the first explicit feature information and the first explicit representation image information into a third training model, wherein the third training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets includes: the first explicit characteristic information, the first extrinsic presentation image information, and an identification result identifying whether the first explicit characteristic information satisfies consistency of the first extrinsic presentation image information;

step 840: obtaining third output information of the third training model, wherein the third output information is an identification result for identifying whether the first explicit feature information meets the first extrinsic representation image information consistency;

step 850: obtaining a second result according to the third output information and the first result, wherein the second result is represented as a screening result of the first endoscope.

Specifically, first explicit characteristic information can be correspondingly obtained according to the subject information, where the first explicit characteristic information is extrinsic representation image information of a human body formed in historical experience of a corresponding detection subject, for example, a human face with a stomach illness may show certain characteristics, such as a black eye, a yellow face, a dark purple mouth, and a thick tongue coating in eyes, so as to obtain current first extrinsic representation image information of a user, and then the first explicit characteristic information and the first extrinsic representation image information are input into a third training model, where an identification result indicating whether the first explicit characteristic information satisfies consistency of the first extrinsic representation image information is obtained through the third training model, so that a third output information of the third training model is obtained, and finally a second result is obtained after comprehensive processing is performed on the third output information and the first result, where the second result is a screening result in the first user, therefore, the accuracy of the screening result of the user is further improved, the doctor is helped to accurately evaluate the illness state of the patient, more scientific treatment selection is provided, and the technical effect of improving the working efficiency of the doctor is achieved.

Further, after obtaining a second result according to the third output information and the first result, step 850 of this embodiment further includes:

step 851: obtaining first body health index information of the first user;

step 852: taking the first body health index information as an abscissa;

step 853: when the second result is healthy, constructing a two-dimensional rectangular coordinate system by taking the second result as a vertical coordinate;

step 854: and constructing a logistic regression line in the two-dimensional rectangular coordinate system according to the logistic regression model to obtain a first risk detection model, wherein one side of the logistic regression line represents a first output result, the other side of the logistic regression line represents a second output result, the first output result indicates that the first user has high risk of contracting a disease, and the second output result indicates that the first user does not have high risk of contracting a disease.

Specifically, the logistic regression model is a machine learning model reflecting the relationship between independent variables and dependent variables, the first body health index information is used as an abscissa, when the screening result of the user is normal monitoring, the subsequent morbidity risk of the user can be predicted, a two-dimensional rectangular coordinate system is constructed by using the second result information as an ordinate, a logistic regression line is obtained through the two-dimensional rectangular coordinate system based on the logistic regression model, one side of the logistic regression line represents a first output result, and the other side of the logistic regression line represents a second output result, for example, the higher the first body health index information index is, the lower the morbidity risk is, the lower the first body health index information index is, the higher the morbidity risk is, the disease probability is judged through the logistic regression line, and the relationship between the first body health index information and the second result of the user is better reflected through the logistic regression model, the risk of future diseases of the user is predicted through the body health index information and the second result information of the user, and then the user can be reminded to prevent in advance according to actual conditions, so that the body health of the user is guaranteed.

Example two

Based on the same inventive concept as that of one endoscope screening processing method in the foregoing embodiments, the present invention also provides an endoscope screening processing method apparatus, as shown in fig. 2, the apparatus including:

a first obtaining unit 11, the first obtaining unit 11 being configured to obtain first image information of an endoscope for a first user;

a second obtaining unit 12, wherein the second obtaining unit 12 is configured to obtain an age of the first user;

a first judging unit 13, wherein the first judging unit 13 is configured to judge whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, obtain first image input data according to the first image information;

a third obtaining unit 14, the third obtaining unit 14 being configured to obtain first age input data according to the first user age;

a first input unit 15, where the first input unit 15 is configured to input the first image input data and the first age input data into a first training model, where the first training model is obtained by training multiple sets of training data, and each set of training data in the multiple sets includes: the first age input data, the first image input data, and identification information identifying whether the first user is normal;

a fourth obtaining unit 16, where the fourth obtaining unit 16 is configured to obtain first output information of the first training model, where the first output information is identification information that identifies whether the first user is normal;

a fifth obtaining unit 17, wherein the fifth obtaining unit 7 is configured to obtain historical diagnosis information of the first user;

a sixth obtaining unit 18, said sixth obtaining unit 18 being configured to obtain a first result from said historical diagnostic information and said first output information, said first result being represented as a screening result of said first user endoscope.

Further, the third obtaining unit further includes:

a seventh obtaining unit for obtaining subject information of the endoscopic examination;

an eighth obtaining unit, configured to obtain first age group information according to the subject information, where the first age group information is probability originating age group information of the subject information;

a ninth obtaining unit configured to obtain first age input data according to the first user age minus the first age group information.

Further, the first determining unit further includes:

a tenth obtaining unit for obtaining first length information of a first user's penetration of the endoscope;

an eleventh obtaining unit, configured to obtain height information of the first user;

a twelfth obtaining unit, configured to obtain a first arrival result according to the first length information and the height information;

a thirteenth obtaining unit configured to obtain first image input data from the first image information when the first arrival result satisfies a second predetermined condition.

Further, the first determining unit further includes:

a fourteenth obtaining unit configured to obtain target image information of a detection position from the subject information;

a second input unit, configured to input the first image and the target image information into a second training model, where the second training model is obtained by training multiple sets of training data, and each set of training data in the multiple sets includes: the first image information, the target image information and an identification result for identifying whether the first image information meets the consistency of the target image information;

a fifteenth obtaining unit, configured to obtain second output information of the second training model, where the second output information identifies whether the first image information satisfies an identification result of target image information consistency;

a sixteenth obtaining unit configured to obtain first image input data from the first image when the second output information is yes.

Further, the first determining unit further includes:

a seventeenth obtaining unit for obtaining second length information of the endoscope going deep into the first user;

an eighteenth obtaining unit for obtaining video stream information of the endoscope in depth to the first user;

a nineteenth obtaining unit configured to obtain N second images according to the video stream information and the second length information;

a second determination unit configured to determine whether or not the N second images are indispensable images of target image information that reaches a detection position;

a twentieth obtaining unit configured to obtain the first image input data from the first image information when the determination result is yes.

Further, the apparatus further comprises:

a twenty-first obtaining unit, configured to obtain first explicit feature information according to the subject information, where the first explicit feature information is extrinsic expression image information of a human body formed in historical experience by the subject information;

a twenty-second obtaining unit configured to obtain first appearance image information;

a third input unit, configured to input the first explicit feature information and the first extrinsic representation image information into a third training model, where the third training model is obtained by training multiple sets of training data, and each set of training data in the multiple sets includes: the first explicit characteristic information, the first extrinsic presentation image information, and an identification result identifying whether the first explicit characteristic information satisfies consistency of the first extrinsic presentation image information;

a twenty-third obtaining unit, configured to obtain third output information of the third training model, where the third output information is an identification result that identifies whether the first explicit feature information satisfies the first extrinsic representation image information consistency;

a twenty-fourth obtaining unit for obtaining a second result from the third output information and the first result, wherein the second result is represented as a screening result of the first user endoscope.

Further, the apparatus further comprises:

a twenty-fifth obtaining unit, configured to obtain first body health index information of the first user;

a first operation unit configured to take the first body health index information as an abscissa;

a second operation unit configured to construct a two-dimensional rectangular coordinate system using the second result as a vertical coordinate when the second result is healthy;

a twenty-sixth obtaining unit, configured to construct a logistic regression line in the two-dimensional rectangular coordinate system according to a logistic regression model, and obtain a first risk detection model, where one side of the logistic regression line represents a first output result, and the other side of the logistic regression line represents a second output result, the first output result indicates that the first user has a high risk probability of contracting a disease, and the second output result indicates that the first user does not have a high risk probability of contracting a disease.

Various modifications and specific examples of an endoscope screening processing method in the first embodiment of fig. 1 are also applicable to an endoscope screening processing device in the present embodiment, and a method for implementing an endoscope screening processing device in the present embodiment is clear to those skilled in the art from the foregoing detailed description of an endoscope screening processing method, so that a detailed description is omitted here for brevity of the description.

EXAMPLE III

Based on the same inventive concept as one of the endoscope screening processing methods in the previous embodiments, the present invention also provides an exemplary electronic device, as shown in fig. 3, including a memory 304, a processor 302, and a computer program stored on the memory 304 and executable on the processor 302, the processor 302 implementing the steps of any one of the endoscope screening processing methods described above when executing the program.

Where in fig. 3 a bus architecture (represented by bus 300), bus 300 may include any number of interconnected buses and bridges, bus 300 linking together various circuits including one or more processors, represented by processor 302, and memory, represented by memory 304. The bus 300 may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. A bus interface 306 provides an interface between the bus 300 and the receiver 301 and transmitter 303. The receiver 301 and the transmitter 303 may be the same element, i.e., a transceiver, providing a means for communicating with various other apparatus over a transmission medium. The processor 302 is responsible for managing the bus 300 and general processing, and the memory 304 may be used for storing data used by the processor 302 in performing operations.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims

1. An endoscopic screening processing apparatus, wherein the apparatus comprises:

a first obtaining unit configured to obtain first image information of an endoscope for a first user;

a second obtaining unit configured to obtain an age of the first user;

a first judging unit configured to judge whether the first image information satisfies a first predetermined condition, and when the first predetermined condition is satisfied, obtain first image input data from the first image information;

a third obtaining unit configured to obtain first age input data according to the first user age;

a first input unit configured to input the first image input data and the first age input data into a first training model, wherein the first training model is obtained by training a plurality of sets of training data, and each set of training data in the plurality of sets includes: the first age input data, the first image input data, and identification information identifying whether the first user is normal;

a fourth obtaining unit, configured to obtain first output information of the first training model, where the first output information is identification information that identifies whether the first user is normal;

a fifth obtaining unit, configured to obtain historical diagnostic information of the first user;

a sixth obtaining unit for obtaining a first result from the historical diagnostic information and the first output information, the first result being represented as a screening result of the first user endoscope;

wherein the first judging unit further includes:

2. The apparatus of claim 1, wherein the third obtaining unit further comprises:

3. The apparatus of claim 1, wherein the first determining unit further comprises:

4. The apparatus of claim 2, wherein the first determining unit further comprises:

5. The apparatus of claim 2, wherein the apparatus further comprises:

6. The apparatus of claim 5, wherein the apparatus further comprises:

7. An endoscope screening process apparatus comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the steps of the apparatus of any one of claims 1-6 are carried out when the program is executed by the processor.