CN114639479A - Intelligent diagnosis auxiliary system based on medical knowledge map - Google Patents
Intelligent diagnosis auxiliary system based on medical knowledge map Download PDFInfo
- Publication number
- CN114639479A CN114639479A CN202210259321.8A CN202210259321A CN114639479A CN 114639479 A CN114639479 A CN 114639479A CN 202210259321 A CN202210259321 A CN 202210259321A CN 114639479 A CN114639479 A CN 114639479A
- Authority
- CN
- China
- Prior art keywords
- module
- information
- treatment
- diagnosis
- disease
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Abstract
An intelligent diagnosis auxiliary system based on medical knowledge maps belongs to the field of medical diagnosis. The system aims to solve the problems that a traditional medical diagnosis system is difficult to integrate information, not only causes a great deal of waste of medical information, but also cannot perform effective auxiliary diagnosis, is low in treatment efficiency, cannot perform effective intelligent triage prediction, easily generates treatment burden and hospital operation burden for patients, is low in working efficiency, cannot give corresponding treatment results timely and quickly, and influences the diagnosis accuracy. The problem of the existing isolated island of medical data is solved, intelligent auxiliary diagnosis of common diseases is realized, the efficiency of seeing a doctor is improved, auxiliary diagnosis with higher reliability is provided for doctors, intelligent triage prediction is realized, the probability of error triage is reduced, efficient and convenient pre-diagnosis consultation is provided for patients, the burden of the patients and the operation load of hospitals are reduced, and the diagnosis accuracy of the comprehensive data analysis diagnosis auxiliary system for the information of the patients is gradually improved through artificial error correction.
Description
Technical Field
The invention relates to the field of medical diagnosis, in particular to an intelligent diagnosis auxiliary system based on a medical knowledge graph.
Background
The medical field is characterized by data-intensive, knowledge-intensive and mental-labor-intensive fields. Massive data continuously generated in the medical industry is mostly unstructured data. With the continuous improvement of the medical system in China, medical resources including medical equipment and medical staff are gradually increased, but the situations of shortage of medical resources and low hospital operating efficiency still exist, the distribution of the medical resources in China is unbalanced, the original disease diagnosis method is often restricted by the actual experience of a decision maker, and is easily interfered by the current subjective consciousness of the decision maker and the external environment, so the diagnosis result is deviated, and the traditional hospital diagnosis system also has the following defects when in use:
1. the problem of medical data isolated island exists in the hospital, information is difficult to integrate, a large amount of waste of medical information is caused, effective auxiliary diagnosis cannot be carried out, and the efficiency of seeing a doctor is low.
2. When the patient is in initial diagnosis, effective intelligent diagnosis forecasting can not be carried out, the burden of seeing a doctor and the operation burden of a hospital are easily generated for the patient, and the working efficiency is low.
3. When a patient is diagnosed with a disease, a corresponding diagnosis result cannot be given timely and quickly, a large amount of examinations are needed, so that not only is a certain economic loss brought to the patient, but also the waste of medical resources is easily caused, and the accuracy of diagnosis is affected.
Disclosure of Invention
The invention aims to provide an intelligent diagnosis auxiliary system based on a medical knowledge graph, which reduces the burden of a doctor for inputting medical records by establishing an electronic medical record of a patient, fuses medical knowledge by establishing the medical knowledge graph, overcomes the problem of the existing isolated medical data, improves the utilization rate of medical data, realizes the intelligent auxiliary diagnosis of common diseases, is convenient for the patient to find characteristic diseases in time when the patient is hospitalized, improves the efficiency of hospitalization, brings convenience for the patient to hospitalize, carries out corresponding diagnosis through an automatic diagnosis module, analyzes to obtain suspected diseases with weight, provides auxiliary diagnosis with higher reliability for the doctor, helps to improve the diagnosis efficiency and the accuracy of the doctor, realizes intelligent triage prediction, reduces the probability of wrong triage, provides efficient and convenient pre-diagnosis consultation for the patient, reduces the burden of the patient and the operation load of a hospital, the system has the advantages that efficient and accurate medical experience is brought to patients, more objective and reliable diagnosis and treatment suggestions compared with the existing artificial intelligent diagnosis system can be provided after more influence factors are considered, the trouble that equipment is insufficient under the complexity of user input information and the environment where the user is located is avoided, the user can make intelligent diagnosis and treatment of diseases at any time and any place as long as the place with a network is provided, the diagnosis accuracy of the comprehensive data analysis and diagnosis auxiliary system for patient information is gradually improved through artificial error correction, and the problems in the background technology are solved.
In order to achieve the purpose, the invention provides the following technical scheme:
the utility model provides an intelligent diagnosis auxiliary system based on medical treatment knowledge map, includes intelligent diagnosis system, be equipped with patient data layer, entity extraction layer, supplementary diagnosis layer and recommended treatment layer in the intelligent diagnosis system, the output on patient data layer is connected with the input on entity extraction layer, and the output on entity extraction layer is connected with the input on supplementary diagnosis layer, and the output on supplementary diagnosis layer is connected with the input on recommended treatment layer.
The patient data layer comprises an information acquisition module, an information storage module and a medical record generation module.
The entity extraction layer comprises an information extraction module, an information identification module, an information output module and a medical knowledge map.
Wherein the auxiliary diagnosis layer comprises an information processing module, an automatic diagnosis module and a recommendation checking module.
The recommended treatment layer comprises a treatment scheme generation module, a personalized recommendation module and an information integration module.
Furthermore, the output end of the information acquisition module is connected with the input end of the information storage module, the output end of the information storage module is connected with the output end of the medical record generation module, and the output end of the medical record generation module is connected with the input end of the entity extraction layer. The information acquisition module is internally provided with a data input module, the data input module adopts manual input or voice input to generate a text for the patient to complain, and the output result of the entity extraction layer consists of disease types, disease symptoms and inducement.
Further, the information acquisition module acquires information including patient ID, gender, age, disease type, medical history, taboo and key living environment, the key living environment is used as an evaluation basis to quantify the probability that the patient suffers from the target disease into three levels, namely extremely easy, common and difficult, and the patient information is a threshold value of the mapping relation between the patient information and the target disease in the medical record learning system and a training set of the initially constructed mapping relation.
Furthermore, the output end of the information extraction module is connected with the input end of the information identification module, the output end of the information identification module is connected with the input end of the information output module, the output end of the information output module is connected with the input end of the medical knowledge map, and the output end of the medical knowledge map is connected with the input end of the auxiliary diagnosis layer.
Furthermore, the information extraction module is used for extracting medical record information of the patient from the patient data layer, the information identification module is used for carrying out diagnosis analysis on the information extracted from the patient data layer, and the analysis result is output to a medical knowledge map through the information output module, wherein the medical knowledge map comprises a disease knowledge base, an examination and examination knowledge base, a symptom knowledge base, a medicine knowledge base, a body part knowledge base and an operation knowledge base.
Further, the medical knowledge graph is based on a triple representation mode and comprises symptom, disease, part, medicine, department and crowd basic entity information, the medical knowledge graph is classified according to gender, and comprises a part symptom relation, a part disease relation, a symptom disease relation, a disease department relation, a medicine disease relation, a medicine symptom relation and a medicine crowd relation, the entity and the relation are extracted by adopting a method based on an entity dictionary, an entity rule and mode matching, the quantification of the symptom and disease relation is realized, and the medical knowledge graph is clear and easy to understand, so that part of secondary information needs to be filtered, main information needs to be extracted, and the results are randomly ordered.
Furthermore, the output end of the information processing module is connected with the input end of the automatic diagnosis module, the output end of the automatic diagnosis module is connected with the input end of the recommended examination module, and the output end of the recommended examination module is connected with the input end of the recommended treatment layer.
Furthermore, the information processing module comprises an information comparison module and an information analysis module, the output end of the information comparison module is connected with the input end of the information analysis module, the information comparison module compares the patient information extracted by the entity extraction layer and analyzes the patient information through the information analysis module, the automatic diagnosis module diagnoses suspected diseases of the patient, and a corresponding recommended examination list is given through the recommended examination module.
Furthermore, the output end of the treatment scheme generation module is connected with the input end of the personalized recommendation module, and the output end of the personalized recommendation module is connected with the input end of the information integration module.
Further, the treatment scheme generation module is used for generating a preliminary treatment scheme according to the result of the examination report and the diagnosis result of the doctor, the personalized recommendation module is used for searching for treatment modes of similar patient groups according to the basic information of the patient and the diagnosis result of the doctor to generate personalized recommendations, and the information integration module is used for integrating the treatment scheme and the personalized recommendations by utilizing a linear model fusion technology to obtain the final treatment scheme for treatment.
Further, the treatment plan generation module generates the treatment plan including the steps of:
obtaining examination results and a final treatment plan of a patient;
checking and analyzing the checking result, and determining the number of items of the checking result and the disease characteristics of each checking result;
dividing treatment sub-schemes of the treatment scheme according to the number of items, adapting the disease characteristics and the treatment sub-schemes, and determining an adaptation value corresponding to the disease characteristics of each inspection result;
using the fitting value as a first treatment standard parameter;
acquiring a recommended treatment scheme through big data according to the examination result of the patient; wherein the content of the first and second substances,
the recommended treatment is not less than 2;
dividing the recommended treatment scheme into recommended treatment sub-schemes according to the items, adapting the disease characteristics and the recommended treatment sub-schemes, and determining a secondary scheme adaptation value corresponding to the disease characteristics of each inspection result; wherein the content of the first and second substances,
the secondary scheme adaptation value corresponding to each disease characteristic is not lower than two;
obtaining a target treatment sub-regimen exceeding the first treatment standard parameter by comparing the secondary regimen adaptation value for each disease feature with the first treatment standard parameter; wherein, the first and the second end of the pipe are connected with each other,
performing a same targeted therapeutic sub-regimen fusion when there are multiple targeted therapeutic sub-regimens per disease feature;
according to the target treatment sub-scheme, carrying out scheme fusion and determining a final diagnosis and treatment scheme; when the schemes are fused, multidimensional fusion is adopted, and each disease feature after fusion is only a unique and determined unique treatment sub-scheme.
In the prior art, for the recommendation of treatment schemes, most of the recommendations of enterprises and systems are based on deep learning networks of big data, and the most suitable scheme is matched, but the big data also has certain defects. That is, a model of big data, in terms of training and raw data, if the sample of raw data is not correct, or there is insufficient training times in the training phase, the training loss is too large, which may cause the recommendation of the treatment plan to be wrong. In hospitals, the trained model is applied, the training stage and the original data of the model cannot be judged to be correct,
further, after the treatment scheme generation module performs scheme fusion, the method also comprises the step of calculating the confidence level of the scheme fusion, and the specific steps are as follows:
step 1: calculating a gray correlation between different target treatment sub-protocols according to the target treatment sub-protocol by:
wherein X (K) represents the treatment characteristics of the Kth treatment sub-regimen; x (m) represents the treatment characteristics of the m-th seed treatment sub-regimen; k is not equal to m, and the m belongs to a positive integer; rho belongs to [ 0-1 ], and rho represents a normalized value of the treatment characteristics;
step 2: calculating fusion expectation, fusion entropy and fusion super-entropy among any target treatment sub-schemes according to the relevance:
wherein n represents the total number of target treatment sub-regimens;
a characteristic mean value representing a target treatment sub-regimen; xiRepresenting a treatment profile of an ith target treatment sub-regimen; q represents the fusion expectation between the target therapeutic sub-regimens; s represents the fusion entropy between the target treatment sub-protocols;
representing a fusion superentropy between target treatment sub-protocols;
and step 3: determining a trust value according to the fusion expectation, the fusion entropy and the fusion super-entropy:
wherein Z represents the confidence level of the solution fusion.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention relates to an intelligent diagnosis auxiliary system based on a medical knowledge graph, which collects the information of a patient through an information collection module, establishes an electronic medical record of the patient, relieves the burden of entering the medical record of a doctor, is beneficial to unified and standard management of medical data, fuses medical knowledge through establishing the medical knowledge graph, overcomes the problem of the existing isolated island of the medical data, improves the utilization rate of the medical data, realizes the intelligent auxiliary diagnosis of common diseases, and realizes the intelligent auxiliary diagnosis of the common diseases, wherein an established characteristic disease database is used for storing and analyzing the characteristic information of the patient suffering from the disease extracted by a system and the possible disease range mapped by the characteristic information, so that the characteristic disease can be found in time when the patient is hospitalized, the hospitalizing efficiency is improved, the convenience is brought to the patient, and the possible diseases of the patient can be analyzed through an information comparison module and an information analysis module, and corresponding diagnosis is carried out through the automatic diagnosis module, so that suspected diseases with weight are obtained through analysis, auxiliary diagnosis with higher reliability is provided for doctors, and the diagnosis efficiency and accuracy of the doctors are improved.
2. According to the intelligent diagnosis auxiliary system based on the medical knowledge map, the department is recommended through the recommendation and inspection module, the list of recommended inspection items is recommended, the intelligent triage prediction is realized, the probability of wrong triage is reduced, efficient and convenient pre-diagnosis consultation is provided for patients, the burden of the patients and the operation load of hospitals are reduced, efficient and accurate medical experience is brought to the patients, more objective and reliable diagnosis and treatment suggestions can be provided compared with the existing artificial intelligent diagnosis and treatment system after more influence factors are considered, the troubles of complexity of information input by users and insufficient equipment in the environment where the users are located are avoided, and the users can perform intelligent diagnosis and treatment of diseases at any place and any time as long as the places with networks exist.
3. The invention relates to an intelligent diagnosis auxiliary system based on a medical knowledge map.A specialist database is used for storing a final output result of a comprehensive data analysis system for artificial intelligent disease diagnosis and establishing a supplementary retrieval tool, a doctor can conveniently review and judge whether a disease process and a result of a patient are accurate through the system by the aid of the establishment of the retrieval tool, the diagnosis accuracy of the comprehensive data analysis diagnosis auxiliary system for patient information is gradually improved through artificial error correction, and the supplementary retrieval tool has the effects of facilitating the investigation and correction of in-doubt diagnosis results by an outpatient doctor and the diagnosis accuracy.
Drawings
FIG. 1 is a schematic diagram of an overall system of an intelligent diagnosis assistance system based on a medical knowledge graph according to the present invention;
FIG. 2 is a main block diagram of an intelligent diagnosis assistance system based on medical knowledge-graph according to the present invention;
FIG. 3 is a schematic diagram of the module connections of an intelligent medical knowledge-based diagnosis assistance system according to the present invention;
FIG. 4 is a schematic diagram showing the connection of some modules of an intelligent medical knowledge-based diagnosis assistance system according to the present invention;
FIG. 5 is a schematic diagram of some modules of an intelligent medical knowledge-map-based diagnostic support system according to the present invention;
fig. 6 is a schematic working flow diagram of an intelligent diagnosis assisting system based on a medical knowledge map according to the present invention.
In the figure: 1. a patient data layer; 11. an information acquisition module; 12. an information storage module; 13. a medical record generation module; 2. an entity extraction layer; 21. an information extraction module; 22. an information identification module; 23. an information output module; 24. a medical knowledge map; 3. auxiliary fault diagnosis; 31. an information processing module; 311. an information comparison module; 312. an information analysis module; 32. an automatic diagnostic module; 33. a recommendation checking module; 4. recommending a treatment layer; 41. a treatment plan generation module; 42. a personalized recommendation module; 43. an information integration module; 5. an intelligent diagnostic system.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
Example one
Referring to fig. 1-4, an intelligent diagnosis assisting system based on medical knowledge graph comprises an intelligent diagnosis system 5, a patient data layer 1 and an entity extraction layer 2 are arranged in the intelligent diagnosis system 5, auxiliary diagnosis layer 3 and recommended treatment layer 4, the output on patient data layer 1 is connected with the input on entity extraction layer 2, the output on entity extraction layer 2 is connected with auxiliary diagnosis layer 3's input, auxiliary diagnosis layer 3's output is connected with the input on recommended treatment layer 4, patient data layer 1 includes information acquisition module 11, information storage module 12 and case history generation module 13, information acquisition module 11's output is connected with information storage module 12's input, information storage module 12's output is connected with case history generation module 13's output, case history generation module 13's output is connected with entity extraction layer 2's input. The information acquisition module 11 is internally provided with a data input module which adopts manual input or voice input to generate a text, the output result of the entity extraction layer 2 consists of disease types, disease symptoms and inducement, the information acquisition module 11 acquires three grades of the ID, sex, age, disease types, disease histories, taboos and important living environments of patients, the important living environments serve as evaluation bases to quantify the probability that the patients have target diseases into three grades of easy, common and difficult, the patient information is a threshold value of the mapping relationship between the patient information and the target diseases in a case history learning system and a training set for initially constructing the mapping relationship, the entity extraction layer 2 comprises an information extraction module 21, an information recognition module 22, an information output module 23 and a medical knowledge map 24, the output end of the information extraction module 21 is connected with the input end of the information recognition module 22, the output end of the information identification module 22 is connected with the input end of the information output module 23, the output end of the information output module 23 is connected with the input end of the medical knowledge map 24, the output end of the medical knowledge map 24 is connected with the input end of the auxiliary diagnosis layer 3, the information extraction module 21 is used for extracting the medical record information of the patient from the patient data layer 1, the information identification module 22 is used for diagnosing and analyzing the information extracted from the patient data layer 1, the analysis result is output to the medical knowledge map 24 through the information output module 23, the medical knowledge map 24 comprises a disease knowledge base, an examination and examination knowledge base, a symptom knowledge base, a medicine knowledge base, a body part knowledge base and an operation knowledge base, the medical knowledge map 24 is based on the representation mode of triples, comprises the basic entity information of symptoms, diseases, parts, medicines, departments and crowds, and is classified according to the sex, and the medical knowledge map 24 must be clear and easy to understand, so that part of secondary information needs to be filtered, main information is extracted, and the results are randomly sorted, the auxiliary diagnosis layer 3 comprises an information processing module 31, an automatic diagnosis module 32 and a recommendation check module 33, the output end of the information processing module 31 is connected with the input end of the automatic diagnosis module 32, the output end of the automatic diagnosis module 32 is connected with the input end of the recommendation check module 33, the output end of the recommendation check module 33 is connected with the input end of the recommendation treatment layer 4, the information processing module 31 comprises an information comparison module 311 and an information analysis module 312, the output end of the information comparison module 311 is connected with the input end of the information analysis module 312, the information comparison module 311 compares the patient information extracted by the entity extraction layer 2, and analyzes the patient information through the information analysis module 312, the automatic diagnosis module 32 diagnoses suspected diseases of the patient, and gives a corresponding recommended examination list through the recommended examination module 33, the recommended treatment layer 4 comprises a treatment plan generation module 41, a personalized recommendation module 42 and an information integration module 43, the output end of the treatment plan generation module 41 is connected with the input end of the personalized recommendation module 42, the output end of the personalized recommendation module 42 is connected with the input end of the information integration module 43, the treatment plan generation module 41 is used for generating a preliminary treatment plan according to the result of the examination report and the diagnosis result of the doctor, the personalized recommendation module 42 is used for searching the treatment mode of similar patient population according to the basic information of the patient and the diagnosis result of the doctor to generate personalized recommendation, the information integration module 43 is used for integrating the treatment plan and the personalized recommendation by using a linear model fusion technology to obtain a final treatment plan treatment.
Please refer to fig. 5-6: an intelligent diagnosis assisting system based on a medical knowledge map comprises the following operation steps:
s1: and establishing patient information and establishing a corresponding electronic medical record according to the patient information.
S101: the information acquisition module 11 acquires the information including the ID, sex, age, disease type, medical history, contraindications and important living environment of the patient, the important living environment is used as evaluation basis to quantify the probability that the patient has the target disease into three grades which are extremely easy, common and difficult, and the information storage module 12 stores the acquired data.
S102: the medical record generation module 13 generates a corresponding electronic medical record for the visit information of the patient.
S2: and (4) building a medical knowledge map frame and analyzing the information of the doctor.
S201: the construction process of the medical knowledge graph 24 sequentially comprises three steps of establishing a knowledge base mode graph, mode graph definition, knowledge extraction and knowledge fusion, wherein the mode graph definition comprises concepts owned by the knowledge base, attributes of the concepts and hierarchical relations among the concepts, the common traditional Chinese medicine knowledge base of the knowledge base mode graph mainly comprises upper-layer concepts of traditional Chinese medicines, traditional Chinese medicine syndromes, traditional Chinese medicine diseases and the like and the attributes of the concepts, the mode graph is constructed, the knowledge extraction mainly comprises network related entities, entity types, synonym relations and attribute value relations of the traditional Chinese medicines, the knowledge fusion is to fuse the content of the knowledge extraction, the medical knowledge base used for constructing the medical knowledge graph 24 needs to identify the related entities of the medicines from professional medicine texts, and a large-scale corpus is used for learning out a labeling model and further labeling sentences.
S202: and establishing a characteristic disease database, wherein the characteristic disease database is used for storing and analyzing the characteristic information of the sick patient extracted by the system and the possible disease range mapped by the characteristic information.
S203: the information extraction module 21 extracts the patient medical record information collected from the information collection module 11, stores the extracted patient information and the possible disease range mapped by the characteristic information in the characteristic disease database, and the information recognition module 22 is used for diagnosing and analyzing the information extracted from the information collection module 11, and the analyzed result is output to the medical knowledge map 24 through the information output module 23.
S3: and performing auxiliary diagnosis on the extracted patient information, and giving a corresponding recommended examination list.
S301: the expert database is established and used for storing the final output result of the artificial intelligence disease diagnosis comprehensive data analysis system and establishing a supplementary retrieval tool, the supplementary retrieval tool is used for facilitating the investigation and correction of the in-doubt diagnosis result by an outpatient doctor, and the expert database is operated and referred by the doctor and is the core of the system reference significance.
S302: the patient information extracted by the entity extraction layer 2 is compared with the information in the medical knowledge map 24 by the information comparison module 311, and the compared information is subjected to data analysis by the information analysis module 312.
S303: the inference engine is used to apply rules to the extracted patient information to infer the information of the patient and diagnose the suspected disease of the patient via the automatic diagnosis module 32.
S303: calling the final output result of the characteristic disease database comprehensive data analysis system in the expert database, establishing a supplementary retrieval tool, facilitating doctors to review and judge whether the disease process and the result of the patients are accurate through the system, gradually improving the diagnosis accuracy rate of the comprehensive data analysis diagnosis auxiliary system through artificial error correction, and giving a corresponding recommended inspection list through a recommended inspection module 33, wherein the recommended inspection list comprises suspected disease types, recommended departments and recommended inspection items.
S4: recommending a treatment plan and generating personalized recommendations.
S401: the treatment plan generation module 41 performs association rule analysis on the clinical guideline, the medical knowledge picture and the clinical data, and generates a preliminary treatment plan according to the result of the examination report and the diagnosis result of the doctor.
S402: and finding out clinically similar patient groups according to the basic information of the patients and the diagnosis results of doctors, and generating personalized recommendations by combining the treatment modes of the similar patient groups and the common treatment modes.
S403: and the information integration module 43 integrates the treatment scheme and the personalized recommendation by using a linear model fusion technology to obtain the final treatment scheme.
To sum up, the intelligent diagnosis auxiliary system based on the medical knowledge graph of the invention acquires the information of the patient through the information acquisition module 11, establishes the electronic medical record of the patient, reduces the burden of the medical record input by the doctor, is beneficial to the unified and standard management of the medical data, fuses the medical knowledge through establishing the medical knowledge graph 24, overcomes the problem of the existing isolated medical data, improves the utilization rate of the medical data, realizes the intelligent auxiliary diagnosis of common diseases, and is convenient for the patient to find characteristic diseases in time when the patient is hospitalized, improves the efficiency of the hospitalization, brings convenience for the patient to hospitalize, analyzes the possible diseases of the patient through the information comparison module 311 and the information analysis module 312, the automatic diagnosis module 32 is used for corresponding diagnosis, suspected diseases with weight are obtained through analysis, auxiliary diagnosis with higher reliability is provided for doctors, the diagnosis efficiency and the accuracy of the doctors are improved, departments and recommended examination item lists are recommended through the recommended examination module 33, intelligent triage prediction is realized, the probability of wrong triage is reduced, efficient and convenient pre-diagnosis consultation is provided for patients, the burden of the patients and the operation load of hospitals are reduced, efficient and accurate medical experience is brought to the patients, the expert database is used for storing the final output result of the comprehensive data analysis system for artificial intelligent disease diagnosis and supplementing the establishment of a retrieval tool, the establishment of the retrieval tool is convenient for doctors to review and judge whether the disease process and the result of the patients are accurate through the system, and the diagnosis accuracy of the comprehensive data analysis diagnosis auxiliary system for patient information is gradually improved through artificial error correction, the supplementary retrieval tool has the advantages that the survey and the correction of in-doubt diagnosis results by outpatient doctors are facilitated, the expert database is used for doctors to operate and refer, the system is the core of system reference significance, more objective and reliable diagnosis and treatment opinions compared with the conventional artificial intelligent diagnosis system can be given after more influence factors are considered, the trouble that the complexity of information input by a user and the equipment of the user is insufficient in the environment is avoided, and the user can carry out intelligent diagnosis and treatment of diseases at any time and any place as long as the user has a network.
Further, the treatment plan generation module generates the treatment plan including the steps of:
acquiring an examination result of a patient and a treatment scheme issued by a doctor;
checking and analyzing the checking result, and determining the number of items of the checking result and the disease characteristics of each checking result;
dividing treatment sub-schemes of the treatment scheme according to the number of items, adapting the disease characteristics and the treatment sub-schemes, and determining an adaptation value corresponding to the disease characteristics of each inspection result;
using the fitting value as a first treatment standard parameter;
acquiring a recommended treatment scheme through big data according to the examination result of the patient; wherein the content of the first and second substances,
the recommended treatment is not less than 2;
performing recommended treatment sub-scheme division on the recommended treatment scheme according to the items, adapting the disease characteristics and the recommended treatment sub-scheme, and determining a secondary scheme adaptation value corresponding to the disease characteristics of each inspection result; wherein, the first and the second end of the pipe are connected with each other,
the secondary scheme adaptation value corresponding to each disease characteristic is not lower than two;
obtaining a target treatment sub-regimen exceeding the first treatment standard parameter by comparing the secondary regimen adaptation value for each disease feature with the first treatment standard parameter; wherein the content of the first and second substances,
performing a same targeted therapeutic sub-regimen fusion when there are multiple targeted therapeutic sub-regimens per disease feature;
according to the target treatment sub-scheme, carrying out scheme fusion and determining a final diagnosis and treatment scheme; when the schemes are fused, multi-dimensional fusion is adopted, and each disease feature after fusion only has a unique determined unique treatment sub-scheme.
In the prior art, for recommendation of treatment schemes, most of recommendations of enterprises and systems are based on deep learning networks of big data, and the most suitable scheme is matched, but the big data also has certain defects. That is, a large data model, and in terms of training and raw data, if the sample of raw data is not correct, or there is insufficient training times in the training stage, the training loss is too large, which may cause the recommendation of a treatment plan to be wrong. In hospitals, the trained model is applied, and the training stage and the original data are not judged to be correct. Therefore, the diagnosis and treatment scheme provided by the doctor and the diagnosis and treatment scheme provided by the big data are compared and fused to generate a unique treatment scheme. In the process of this fusion, the present invention has the following steps: first, each therapeutic purpose of the disease is evaluated according to the technical solutions presented by the doctor, for example: when people treat cold, some medicines are used for reducing fever, which is a sub-scheme, and some medicines are used for treating cough, which is a sub-scheme; each treatment objective corresponds to each disease characteristic, and in this way, the adaptive value of each treatment objective of the treatment scheme facing each disease characteristic treatment can be determined; the doctor's plan is used as a standard, then, the doctor determines a treatment plan through big data, if the treatment effect is to exceed the adaptive value, the treatment plan is better than the doctor's treatment plan, at this time, the treatment plan can be replaced, but if a plurality of treatment plans are better than the plan given by the doctor, the multi-dimensional fusion can be adopted, namely, the advantage of each treatment sub-plan is coordinated, and the treatment sub-plan corresponding to each disease characteristic is given, namely, the fusion of the treatment plan is carried out. Finally we determine the final treatment after fusion of all sub-protocols. The scheme is based on the scheme issued by a doctor, so that the problem of large data deviation is solved without overlarge deviation. Moreover, the problem of insufficient training times can be solved only by fusing a more excellent scheme.
Further, after the treatment plan generating module 41 performs plan fusion, it further calculates the confidence of the plan fusion, and the specific steps are as follows:
step 1: calculating a gray correlation between different target treatment sub-protocols according to the target treatment sub-protocol by:
wherein X (K) represents the treatment characteristics of the Kth treatment sub-regimen; x (m) represents the therapeutic characteristics of the m-th sought therapeutic sub-regimen; k is not equal to m, and the m belongs to a positive integer; rho belongs to [ 0-1 ], and rho represents a normalized value of the treatment characteristics;
the gray relevance degree is calculated by the method as long as the gray relevance degree is calculated through the correlation among different target treatment sub-schemes, the gray relevance degree is the correlation among the different target treatment sub-schemes, and [ X (K) and X (m) ] represent the different treatment sub-schemes, and the calculation based on the normalized value is mainly used for ensuring that the relevance degree among the schemes can be (0-1), and the relevance degree cannot be one hundred percent.
Step 2: calculating fusion expectation, fusion entropy and fusion super-entropy among any target treatment sub-schemes according to the relevance:
wherein n represents the total number of target therapeutic sub-regimens;
a characteristic mean value representing a target treatment sub-regimen; xiRepresenting a treatment characteristic of the ith target treatment sub-regimen; q represents the fusion expectation between the target therapeutic sub-regimens; s represents the fusion entropy between the target treatment sub-protocols;
representing a fusion superentropy between target treatment sub-protocols;
the role of the above three values is mainly to determine the effect of each treatment sub-regimen:
firstly, fusion expectation is obtained, and a fused result is determined;
the fusion entropy reflects the chaos degree after fusion, and the treatment effect is reversely determined.
Finally, the super entropy is used for judging the degree of positive and negative deviation after the fusion of a plurality of schemes, namely after the fusion, the result that the fusion cannot be carried out is an error treatment scheme, and the deviation cannot be caused from the normal treatment scheme.
And step 3: determining a trust value according to the fusion expectation, the fusion entropy and the fusion super-entropy:
wherein Z represents the confidence level of the solution fusion.
From the three values, we determine the final confidence value based on the exponential function, and the confidence value is for each treatment regimen to produce a beneficial therapeutic effect.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be able to cover the technical solutions and the inventive concepts of the present invention within the technical scope of the present invention.
Claims (10)
1. An intelligent diagnosis auxiliary system based on a medical knowledge graph comprises an intelligent diagnosis system (5) and is characterized in that a patient data layer (1), an entity extraction layer (2), an auxiliary diagnosis layer (3) and a recommended treatment layer (4) are arranged in the intelligent diagnosis system (5), the output end of the patient data layer (1) is connected with the input end of the entity extraction layer (2), the output end of the entity extraction layer (2) is connected with the input end of the auxiliary diagnosis layer (3), and the output end of the auxiliary diagnosis layer (3) is connected with the input end of the recommended treatment layer (4);
the patient data layer (1) comprises an information acquisition module (11), an information storage module (12) and a medical record generation module (13);
the entity extraction layer (2) comprises an information extraction module (21), an information identification module (22), an information output module (23) and a medical knowledge map (24);
wherein the auxiliary diagnostic layer (3) comprises an information processing module (31), an automatic diagnosis module (32) and a recommended examination module (33);
the recommended treatment layer (4) comprises a treatment scheme generation module (41), a personalized recommendation module (42) and an information integration module (43).
2. The medical knowledge graph-based intelligent diagnosis assisting system as claimed in claim 1, wherein an output end of the information acquisition module (11) is connected with an input end of the information storage module (12), an output end of the information storage module (12) is connected with an output end of the medical record generation module (13), and an output end of the medical record generation module (13) is connected with an input end of the entity extraction layer (2). The information acquisition module (11) is internally provided with a data input module, the data input module adopts manual input or voice input to generate a text for patient complaints, and the output result of the entity extraction layer (2) consists of disease types, disease symptoms and causes.
3. The medical knowledge graph-based intelligent diagnosis assistance system as claimed in claim 1, wherein the information acquisition module (11) acquires information including patient ID, gender, age, disease type, medical history, contraindications, and important living environment, the important living environment is used as evaluation basis to quantify probability of the patient having the target disease to be extremely easy, and the patient information is "threshold" of mapping relationship between the patient information and the target disease in the case history learning system and training set of mapping relationship initially constructed.
4. The medical knowledge graph-based intelligent diagnosis assistance system as claimed in claim 1, wherein an output end of the information extraction module (21) is connected with an input end of the information identification module (22), an output end of the information identification module (22) is connected with an input end of the information output module (23), an output end of the information output module (23) is connected with an input end of the medical knowledge graph (24), and an output end of the medical knowledge graph (24) is connected with an input end of the auxiliary diagnosis layer (3).
5. The system as claimed in claim 4, wherein the information extraction module (21) is used to extract the patient's medical record information from the patient data layer (1), the information recognition module (22) is used to perform diagnostic analysis on the information extracted from the patient data layer (1), the analysis result is outputted to the medical knowledge map (24) through the information output module (23), the medical knowledge map (24) comprises a disease knowledge base, an examination and examination knowledge base, a symptom knowledge base, a drug knowledge base, a body part knowledge base and an operation knowledge base, the medical knowledge map (24) is based on the representation of triples, including the basic entity information of symptoms, diseases, parts, drugs, departments and groups, and is classified according to the gender, and comprises the part symptom relationship, the part disease relationship, The extraction of the symptom and disease relationship, the disease department relationship, the medicine and disease relationship, the medicine symptom relationship and the medicine crowd relationship adopts a method based on an entity dictionary, an entity rule and pattern matching, the quantification of the symptom and disease relationship is realized, and the medical knowledge graph (24) is clear and easy to understand, so that part of secondary information needs to be filtered, the main information is extracted, and the results are randomly sequenced.
6. The intelligent diagnosis assistance system based on medical knowledge-graph according to claim 1, wherein the output end of the information processing module (31) is connected with the input end of the automatic diagnosis module (32), the output end of the automatic diagnosis module (32) is connected with the input end of the recommended examination module (33), the output end of the recommended examination module (33) is connected with the input end of the recommended treatment layer (4), the information processing module (31) comprises an information comparison module (311) and an information analysis module (312), the output end of the information comparison module (311) is connected with the input end of the information analysis module (312), the information comparison module (311) compares the patient information extracted by the entity extraction layer (2) and analyzes the patient information through the information analysis module (312), and the automatic diagnosis module (32) diagnoses the suspected diseases of the patients, and a corresponding recommended check list is given by a recommended check module (33).
7. The medical knowledge graph-based intelligent diagnosis assistance system as claimed in claim 1, wherein an output end of the treatment plan generation module (41) is connected with an input end of the personalized recommendation module (42), and an output end of the personalized recommendation module (42) is connected with an input end of the information integration module (43).
8. The medical knowledge graph-based intelligent diagnosis assistance system as claimed in claim 7, wherein the treatment plan generating module (41) is configured to generate a preliminary treatment plan according to the result of the examination report and the diagnosis result of the doctor, the personalized recommendation module (42) is configured to search the treatment patterns of similar patient populations according to the basic information of the patient and the diagnosis result of the doctor to generate personalized recommendations, and the information integrating module (43) is configured to integrate the treatment plan and the personalized recommendations to obtain the final treatment plan treatment by using a linear model fusion technique.
9. The medical knowledge-graph-based intelligent diagnosis assistance system according to claim 8, wherein the treatment plan generating module (41) generates the treatment plan including the steps of:
acquiring an examination result of a patient and a treatment scheme of a doctor;
checking and analyzing the checking result, and determining the number of items of the checking result and the disease characteristics of each checking result;
dividing treatment sub-schemes of the treatment scheme according to the number of items, adapting the disease characteristics and the treatment sub-schemes, and determining an adaptation value corresponding to the disease characteristics of each inspection result;
using the fitting value as a first treatment standard parameter;
acquiring a recommended treatment scheme through big data according to the examination result of the patient; wherein the content of the first and second substances,
the recommended treatment is not less than 2;
performing recommended treatment sub-scheme division on the recommended treatment scheme according to the items, adapting the disease characteristics and the recommended treatment sub-scheme, and determining a secondary scheme adaptation value corresponding to the disease characteristics of each inspection result; wherein the content of the first and second substances,
the secondary scheme adaptation value corresponding to each disease characteristic is not lower than two;
obtaining a target treatment sub-regimen exceeding the first treatment standard parameter by comparing the secondary regimen adaptation value for each disease feature with the first treatment standard parameter; wherein, the first and the second end of the pipe are connected with each other,
performing a same targeted therapeutic sub-regimen fusion when there are multiple targeted therapeutic sub-regimens per disease feature;
performing scheme fusion according to the target treatment sub-scheme to determine a final diagnosis and treatment scheme; when the schemes are fused, multidimensional fusion is adopted, and each disease feature after fusion is only a unique and determined unique treatment sub-scheme.
10. The medical knowledge graph-based intelligent diagnosis assistance system according to claim 8, wherein after the treatment plan generation module (41) performs plan fusion, the method further comprises calculating the confidence level of the plan fusion, and the method comprises the following specific steps:
step 1: calculating a gray correlation between different target treatment sub-protocols according to the target treatment sub-protocol by:
wherein X (K) represents the treatment characteristics of the Kth treatment sub-regimen; x (m) represents the treatment characteristics of the m-th seed treatment sub-regimen; k is not equal to m, and the m belongs to a positive integer; rho belongs to [ 0-1 ], and rho represents a normalized value of the treatment characteristics;
step 2: calculating fusion expectation, fusion entropy and fusion super-entropy among any target treatment sub-schemes according to the relevance:
wherein n represents the total number of target treatment sub-regimens;
a characteristic mean value representing a target treatment sub-regimen; xiIndicates the ith target therapeutic subA therapeutic characteristic of the protocol; q represents the fusion expectation between the target therapeutic sub-regimens; s represents the fusion entropy between the target treatment sub-protocols;
representing a fusion superentropy between target treatment sub-protocols;
and step 3: determining a trust value according to the fusion expectation, the fusion entropy and the fusion super-entropy:
wherein Z represents the confidence level of the solution fusion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210259321.8A CN114639479A (en) | 2022-03-16 | 2022-03-16 | Intelligent diagnosis auxiliary system based on medical knowledge map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210259321.8A CN114639479A (en) | 2022-03-16 | 2022-03-16 | Intelligent diagnosis auxiliary system based on medical knowledge map |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114639479A true CN114639479A (en) | 2022-06-17 |
Family
ID=81949128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210259321.8A Pending CN114639479A (en) | 2022-03-16 | 2022-03-16 | Intelligent diagnosis auxiliary system based on medical knowledge map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114639479A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115036034A (en) * | 2022-08-11 | 2022-09-09 | 之江实验室 | Similar patient identification method and system based on patient characterization map |
| CN116246749A (en) * | 2023-05-11 | 2023-06-09 | 西南医科大学附属医院 | Endocrine patient personalized health management system integrating electronic medical records |
| CN116364267A (en) * | 2023-02-16 | 2023-06-30 | 江苏省计算机技术服务有限公司 | Single sign-on management system based on intelligent medical cloud service |
| CN117012374A (en) * | 2023-10-07 | 2023-11-07 | 之江实验室 | Medical follow-up system and method integrating event map and deep reinforcement learning |
| CN117393156A (en) * | 2023-12-12 | 2024-01-12 | 珠海灏睿科技有限公司 | Multi-dimensional remote auscultation and diagnosis intelligent system based on cloud computing |
| CN117690549A (en) * | 2024-02-01 | 2024-03-12 | 中国中医科学院中医临床基础医学研究所 | Traditional Chinese medicine individuation intelligent prescription medicine recommendation system based on similar patient matching |
| CN115954094B (en) * | 2022-12-28 | 2024-05-17 | 广州希华通讯设备有限公司 | Medical waste whole-course traceability intelligent sanitation supervision and management method |
-
2022
- 2022-03-16 CN CN202210259321.8A patent/CN114639479A/en active Pending
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115036034A (en) * | 2022-08-11 | 2022-09-09 | 之江实验室 | Similar patient identification method and system based on patient characterization map |
| CN115036034B (en) * | 2022-08-11 | 2022-11-08 | 之江实验室 | Similar patient identification method and system based on patient characterization map |
| CN115954094B (en) * | 2022-12-28 | 2024-05-17 | 广州希华通讯设备有限公司 | Medical waste whole-course traceability intelligent sanitation supervision and management method |
| CN116364267A (en) * | 2023-02-16 | 2023-06-30 | 江苏省计算机技术服务有限公司 | Single sign-on management system based on intelligent medical cloud service |
| CN116364267B (en) * | 2023-02-16 | 2023-10-31 | 江苏省计算机技术服务有限公司 | Single sign-on management system based on intelligent medical cloud service |
| CN116246749A (en) * | 2023-05-11 | 2023-06-09 | 西南医科大学附属医院 | Endocrine patient personalized health management system integrating electronic medical records |
| CN117012374A (en) * | 2023-10-07 | 2023-11-07 | 之江实验室 | Medical follow-up system and method integrating event map and deep reinforcement learning |
| CN117012374B (en) * | 2023-10-07 | 2024-01-26 | 之江实验室 | Medical follow-up system and method integrating event map and deep reinforcement learning |
| CN117393156A (en) * | 2023-12-12 | 2024-01-12 | 珠海灏睿科技有限公司 | Multi-dimensional remote auscultation and diagnosis intelligent system based on cloud computing |
| CN117393156B (en) * | 2023-12-12 | 2024-04-05 | 珠海灏睿科技有限公司 | Multi-dimensional remote auscultation and diagnosis intelligent system based on cloud computing |
| CN117690549A (en) * | 2024-02-01 | 2024-03-12 | 中国中医科学院中医临床基础医学研究所 | Traditional Chinese medicine individuation intelligent prescription medicine recommendation system based on similar patient matching |
| CN117690549B (en) * | 2024-02-01 | 2024-05-17 | 中国中医科学院中医临床基础医学研究所 | Traditional Chinese medicine individuation intelligent prescription medicine recommendation system based on similar patient matching |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114639479A (en) | Intelligent diagnosis auxiliary system based on medical knowledge map | |
| Lee et al. | Data mining techniques applied to medical information | |
| Wang et al. | Risk assessment of coronary heart disease based on cloud-random forest | |
| CN110111884B (en) | CMKMC-based man-machine cooperative intelligent medical aid decision-making system | |
| Fang et al. | Feature Selection Method Based on Class Discriminative Degree for Intelligent Medical Diagnosis. | |
| CN109378066A (en) | A kind of control method and control device for realizing disease forecasting based on feature vector | |
| CN112164460A (en) | Intelligent disease auxiliary diagnosis system based on medical knowledge map | |
| CN112233736B (en) | Knowledge base construction method and system | |
| CN111191048A (en) | Emergency call question-answering system construction method based on knowledge graph | |
| Folino et al. | A comorbidity-based recommendation engine for disease prediction | |
| CN111243748A (en) | Needle pushing health data standardization system | |
| CN115497631A (en) | Clinical scientific research big data analysis system | |
| CN116779190B (en) | Medical platform user follow-up management system and method based on Internet of things | |
| CN109147954A (en) | The patient information processing unit of knowledge based map | |
| CN113838577A (en) | Convenient layered old people MODS early death risk assessment model, device and establishment method | |
| CN109192312B (en) | Intelligent management system and method for adverse events of heart failure patients | |
| Saleem Durai et al. | An intelligent knowledge mining model for kidney cancer using rough set theory | |
| WO2022141925A1 (en) | Intelligent medical service system and method, and storage medium | |
| CN117457192A (en) | Intelligent remote diagnosis method and system | |
| CN116597981A (en) | Method for preventing, controlling and managing chronic obstructive pulmonary disease by combining digital twin technology | |
| US20090156947A1 (en) | Knowledgebased image informatics system and method | |
| CN113642669B (en) | Feature analysis-based fraud prevention detection method, device, equipment and storage medium | |
| CN113948171A (en) | Matching system for automatically matching hospital grade according to patient condition | |
| AU2021102832A4 (en) | System & method for automatic health prediction using fuzzy based machine learning | |
| CN110189829A (en) | A kind of great chronic disease methods of risk assessment of electric power professional population and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |