CN111383760A

Movatterモバイル変換

Info

Publication number: CN111383760A
Application number: CN202010281330.8A
Authority: CN
Inventors: 韩琳; 李春生
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-04-10
Filing date: 2020-04-10
Publication date: 2020-07-07

Abstract

The invention provides a method for establishing a medical intelligent diagnosis system for nervous system diseases, which comprises the following steps: (1) establishing an artificial neural network model; (2) establishing a knowledge base; (3) classifying and sorting expert knowledge; (4) calculating and outputting by a neural network; (5) matching reasoning to obtain a diagnosis conclusion; (6) if the exact conclusion cannot be obtained, only whether the disease is the nervous system disease or not can be suspected, and the auxiliary examination which should be done in the related knowledge base can be prompted to the user; and (4) when the result of the auxiliary examination is obtained, adding the result into the input information, repeating the steps (3) to (5), and finally obtaining a positive diagnosis result. The invention screens the disease early and intervenes early through big data, finds the disease at the earliest stage in the physical examination process, and diagnoses according to the big database and graph recognition.

Description

Method for establishing medical intelligent diagnosis system for nervous system diseases

Technical Field

The invention belongs to the technical field of intelligent diagnosis, and particularly relates to a medical intelligent diagnosis system establishment method for nervous system diseases.

Background

The diagnosis method commonly used in medicine is to analyze and judge the state of an illness by a doctor according to the examination results of various medical instruments on a patient and by means of the pathological knowledge mastered by the doctor and the accumulated experience for many years, thereby obtaining the diagnosis result. Since the diagnostic method is greatly influenced by subjective factors, the correctness of the diagnostic result is closely related to the medical level of a doctor. Particularly, in hospitals in middle, small and remote areas, the number of experts is small, so that misdiagnosis rates of various difficult and complicated diseases are always high, and a lot of pain and regrets are brought to patients and families. Therefore, people always hope to have an intelligent diagnosis method which can eliminate various human factors and obtain accurate and objective diagnosis results.

The current physical examination results are only subjected to preliminary analysis and are not integrated with daily physical sign indexes and clinical information of customers, so that the results float on the surface, and many people spend expensive detection cost but do not achieve the effect of early intervention. The existing data is inaccurate due to single dimension analysis, and has no expected effect on future life and disease prevention, so that the resource is greatly wasted. In particular, damage to nerve cells is irreversible, and early detection, early diagnosis, and early intervention of nervous system diseases are critical to the effect of treatment.

In the prior art, people usually hope to solve the problems which are difficult to solve in theory but need to be solved urgently in practice through the experience of experts. Medical diagnostic systems of the "expert system" type have been produced. The system stores the experience and knowledge of experts in a computer in a regular form, establishes a knowledge base and carries out medical diagnosis in a symbolic reasoning mode. However, this "duck-fill" knowledge acquisition has met with significant difficulties. On the one hand, the complexity of some difficult conditions makes it difficult to describe them with some rules, even expressed in simple languages, due to the limitations of the knowledge expression of symbols; on the other hand, the rule-based expert system may cause combination explosion due to the sharp increase of the search space with the increase of the scale of the rule base, and a great deal of system time is wasted and the reasoning efficiency is low because the reasoning loop process includes a great number of invalid matching attempts. The fundamental reason for this is that the generative structure and the serial operation of the system have certain drawbacks. Therefore, a medical diagnostic system of a simple "expert system" type can only be used for relatively simple diagnosis of diseases, and is of little value.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a method for establishing a medical intelligent diagnosis system for nervous system diseases, which can accurately diagnose the human brain diseases through a computer and provide an optimal early screening and treatment scheme, and can replace doctors with artificial intelligence in the aspect of nervous system disease diagnosis.

In order to achieve the purpose, the technical scheme of the invention is as follows:

a medical intelligent diagnosis system establishment method for nervous system diseases comprises the following steps:

(1) establishing an artificial neural network model;

(2) establishing a knowledge base: acquiring the clinical experience of a neurology disease expert and knowledge obtained from the Weiwei books and publications;

(3) classifying and sorting expert knowledge: classifying and sorting the information acquired from the interface part of the human-computer interaction terminal according to the knowledge of experts as the input of a neural network;

(4) the neural network calculates and outputs: inputting the input content of the neural network into the network, and immediately obtaining the output result of the network;

(5) matching reasoning to obtain a diagnosis conclusion: matching the output result of the neural network, namely the initial diagnosis result of the system with corresponding expert knowledge, and determining diagnosis if the output result is consistent with the initial diagnosis result of the system;

(6) if the exact conclusion cannot be obtained, only whether the disease is the nervous system disease or not can be suspected, and the auxiliary examination which should be done in the related knowledge base can be prompted to the user; and (4) when the result of the auxiliary examination is obtained, adding the result into the input information, repeating the steps (3) to (5), and finally obtaining a positive diagnosis result.

Preferably, a multi-layer feedforward network is adopted in the step (1), and the training algorithm is an improved single-parameter dynamic search algorithm, namely an SPDS algorithm.

Furthermore, the human-computer interaction machine end comprises a device main body and a writing pen, wherein a writing screen is arranged on the upper surface of the device main body, and a paper inlet and a paper outlet are formed in two corresponding side surfaces of the device main body.

Has the advantages that: the invention screens the disease early and intervenes early through big data, finds the disease at the earliest stage in the physical examination process, and diagnoses according to the big database and graph recognition.

Drawings

Fig. 1 is a schematic structural diagram of an interactive end of the present invention.

In the figure, 1-a device body, 2-writing pens, 2, 3-a writing screen, 4-a paper inlet and 5-a paper outlet.

Detailed Description

The invention is illustrated below with reference to specific examples. It will be understood by those skilled in the art that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention in any way.

(1) establishing an artificial neural network model, adopting a multi-layer feedforward network which is most widely applied at present, and taking an improved single-parameter dynamic search algorithm, namely an SPDS algorithm, as a training algorithm;

As shown in fig. 1, the human-computer terminal includes: the device comprises adevice body 1 and awriting pen 2, wherein a writing screen 3 is arranged on the upper surface of thedevice body 1, and the device body is in the shape of a cuboid box, and apaper inlet 4 and apaper outlet 5 are arranged on two corresponding side faces of the cuboid box.

The working principle of the invention is as follows:

(1) collecting corresponding image information to be diagnosed;

(2) identifying a disease in the nervous system to be diagnosed from the image information;

(3) displaying pre-stored data and a disease image to be diagnosed to a user according to a disease to be diagnosed;

(4) receiving a disease image selected by a user; determining the disease type according to the disease image;

(5) and acquiring treatment measures corresponding to the disease types, and finally printing a report.

Claims

1. A method for establishing a medical intelligent diagnosis system for nervous system diseases is characterized by comprising the following steps:

(1) establishing an artificial neural network model;

2. The method for establishing a medical intelligent diagnosis system for nervous system diseases as claimed in claim 1, wherein in step (1), a multi-layer feedforward network is adopted, and the training algorithm is an improved single parameter dynamic search algorithm, i.e. an SPDS algorithm.

3. The method for establishing a medical intelligent diagnosis system for nervous system diseases as claimed in claim 1, wherein the human-interactive machine end comprises a device body and a writing pen, the upper surface of the device body is provided with a writing screen, and the device body is provided with a paper inlet and a paper outlet at two corresponding sides.