CN112365974A - Learning implementation method and system of clinical assistant decision support system - Google Patents
Learning implementation method and system of clinical assistant decision support system Download PDFInfo
- Publication number
- CN112365974A CN112365974A CN202011231368.0A CN202011231368A CN112365974A CN 112365974 A CN112365974 A CN 112365974A CN 202011231368 A CN202011231368 A CN 202011231368A CN 112365974 A CN112365974 A CN 112365974A
- Authority
- CN
- China
- Prior art keywords
- data
- medical
- medical knowledge
- clinical
- disease diagnosis
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Medical Treatment And Welfare Office Work (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention provides a learning realization method and a system of a clinical assistant decision support system, wherein the method comprises the following steps: acquiring new disease diagnosis data and medical knowledge data through a standardized interface, and counting the newly added sampling number of the disease diagnosis data; when the newly added sampling number reaches a preset data volume, classifying the disease diagnosis data into a medical training data set and a medical verification data set; and updating the clinical assistant model according to the medical training data set and the medical verification data set, and updating the medical knowledge map according to the medical knowledge data to complete the learning and updating of the clinical assistant decision support system. Medical sampling data in HIS of each hospital are mutually communicated, so that a clinical assistant decision support system can learn and update in time, CDSS is updated, the diagnosis capability of doctors is improved, and the misdiagnosis rate is greatly reduced.
Description
Technical Field
The invention relates to the technical field of clinical decision, in particular to a learning implementation method and a learning implementation system of a clinical assistant decision support system.
Background
In recent years, as the degree of science and technology and informatization of the medical industry is gradually increased, a clinical assistance decision support system (CDSS) has attracted more and more attention. The clinical assistant decision support system is a medical information technology application system based on man-machine interaction, and aims to provide clinical decision support for doctors and other health practitioners and complete clinical decision through assistance of data, models and the like. In general, the CDSS presents the knowledge that the doctor cannot remember the knowledge through a computer interface to help the doctor to see the disease. The frequency and convenience of electronic systems is far from comparable to paper materials, which is also the reason why CDSS is now popular in the global medical field.
The conventional CDSS is generally deeply bound with a hospital information management system (HIS) or is a separate mobile phone application. The deeply bound CDSS is difficult to upgrade, HISs of hospitals are not communicated with each other, the later development cost is extremely high, and the application and popularization of the CDSS are not facilitated. The independent mobile phone software CDSS can carry out auxiliary diagnosis and decision after inputting a large amount of information of a patient.
Disclosure of Invention
The invention aims to provide a learning implementation method and a learning implementation system of a clinical assistant decision support system, which are used for realizing mutual communication of medical sampling data in HISs of hospitals, so that the clinical assistant decision support system can learn and update in time, further CDSS is updated, the diagnosis capability of doctors is improved, and the misdiagnosis rate is greatly reduced.
The technical scheme provided by the invention is as follows:
the invention provides a learning implementation method of a clinical assistant decision support system, which comprises the following steps:
acquiring new disease diagnosis data and medical knowledge data through a standardized interface, and counting the newly added sampling number of the disease diagnosis data;
classifying the disease diagnosis data into a medical training data set and a medical verification data set when the newly added sampling number reaches a preset data volume;
and updating a clinical assistant model according to the medical training data set and the medical verification data set, and updating a medical knowledge map according to the medical knowledge data to finish the learning and updating of the clinical assistant decision support system.
Further, the step of obtaining new disease diagnosis data and medical knowledge data through the standardized interface and counting the newly added sampling number of the disease diagnosis data includes:
acquiring disease diagnosis data and medical knowledge data through a standardized interface;
structuring the disease diagnosis data and the medical knowledge data to conform to a preset data structure;
and generating corresponding index identification for each piece of disease diagnosis data, and counting the number of the index identifications to obtain the newly added sampling number of the medical sampling data.
Further, the structuring the disease diagnosis data and the medical knowledge data to conform to a preset data structure comprises the steps of:
judging whether the disease diagnosis data conforms to medical knowledge according to a medical knowledge base;
structuring the medical knowledge data and the disease diagnosis data conforming to medical knowledge to conform to a preset data structure;
and (4) checking the data which accord with the preset data structure and deleting repeated data to obtain final disease diagnosis data and medical knowledge data.
Further, the updating the clinical assistant model according to the medical training data set and the medical verification data set, and the updating the medical knowledge map according to the medical knowledge data to complete the learning update of the clinical assistant decision support system includes the steps of:
inputting the medical training data set into an original clinical auxiliary model for training;
adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
and updating the medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data, and finishing the learning and updating of the clinical assistant decision support system.
Further, the method also comprises the following steps:
inputting the obtained inquiry data into the clinical auxiliary model, and obtaining a probability value of each disease type;
and selecting a preset number of disease types with the probability value arranged in front according to the probability value, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
The invention also provides a learning implementation system of the clinical assistant decision support system, which comprises:
the acquisition module is used for acquiring new disease diagnosis data and medical knowledge data through the standardized interface and counting the newly added sampling number of the disease diagnosis data;
the acquisition module is used for classifying the disease diagnosis data into a medical training data set and a medical verification data set when the newly increased sampling number reaches a preset data volume;
and the learning module is used for updating a clinical auxiliary model according to the medical training data set and the medical verification data set and updating a medical knowledge map according to the medical knowledge data so as to complete the learning and updating of the clinical auxiliary decision support system.
Further, the acquisition module comprises:
the acquisition unit is used for acquiring disease diagnosis data and medical knowledge data through a standardized interface;
the data processing unit is used for carrying out structural processing on the disease diagnosis data and the medical knowledge data so as to accord with a preset data structure;
and the generation counting unit is used for generating corresponding index identification for each piece of disease diagnosis data, and counting the number of the index identifications to obtain the newly added sampling number of the medical sampling data.
Further, the data processing unit includes:
the judging subunit is used for judging whether the disease diagnosis data conforms to medical knowledge or not according to a medical knowledge base;
the processing subunit is used for carrying out structural processing on the medical knowledge data and the disease diagnosis data conforming to the medical knowledge so as to conform to a preset data structure;
and the screening and removing subunit is used for carrying out duplicate checking on the data which accord with the preset data structure and deleting the duplicate data to obtain final disease diagnosis data and medical knowledge data.
Further, the learning module includes:
the training unit is used for inputting the medical training data set into an original clinical auxiliary model for training;
the verification updating unit is used for adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
and the map updating unit is used for updating the medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data and finishing the learning and updating of the clinical assistant decision support system.
Further, the method also comprises the following steps:
the input module is used for acquiring inquiry data;
the processing module is used for inputting the acquired inquiry data into the clinical auxiliary model and acquiring the probability value of each disease type; and selecting a preset number of disease types with the probability values arranged in the front according to the probability values, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
By the learning implementation method and the system of the clinical assistant decision support system, medical sampling data in HIS of each hospital can be communicated with each other, so that the clinical assistant decision support system can be timely learned and updated, CDSS is updated, the diagnosis capability of doctors is improved, and the misdiagnosis rate is greatly reduced.
Drawings
The above features, technical features, advantages and modes of implementation of a learning implementation method and system for a clinical assistant decision support system will be further described in the following detailed description of preferred embodiments in a clearly understandable manner with reference to the accompanying drawings.
FIG. 1 is a flow chart of one embodiment of a learning implementation of a clinical assistance decision support system of the present invention;
FIG. 2 is a flow chart of another embodiment of a learning implementation of a clinical assistance decision support system of the present invention;
FIG. 3 is a flow chart of another embodiment of a learning implementation of a clinical assistance decision support system of the present invention;
FIG. 4 is a flow chart of another embodiment of a learning implementation of a clinical assistance decision support system of the present invention;
FIG. 5 is a schematic structural diagram of an embodiment of a learning implementation system of a clinical assistant decision support system of the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
For the sake of simplicity, the drawings only schematically show the parts relevant to the present invention, and they do not represent the actual structure as a product. In addition, in order to make the drawings concise and understandable, components having the same structure or function in some of the drawings are only schematically illustrated or only labeled. In this document, "one" means not only "only one" but also a case of "more than one".
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
In addition, in the description of the present application, the terms "first", "second", and the like are used only for distinguishing the description, and are not intended to indicate or imply relative importance.
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will illustrate specific embodiments of the present invention with reference to the drawings. It is obvious that the drawings in the following description are only some examples of the invention, and that for a person skilled in the art, other drawings and embodiments can be derived from them without inventive effort.
In an embodiment of the present invention, as shown in fig. 1, a learning implementation method of a clinical assistant decision support system includes:
s100, acquiring new disease diagnosis data and medical knowledge data through a standardized interface, and counting the newly added sampling number of the disease diagnosis data;
specifically, standardized interfaces include, but are not limited to, a database interface, an HL7 interface, and a Web Service interface. The clinical assistant decision support system (CDSS) expands a plurality of standardized interfaces, is connected with a plurality of data sources through the standardized interfaces so as to acquire new medical sampling data, and counts the number of newly acquired disease diagnosis data to obtain the number of newly added samples.
S200, when the newly added sampling number reaches a preset data volume, classifying the disease diagnosis data into a medical training data set and a medical verification data set;
specifically, the CDSS system collects new medical sampling data in real time, periodically determines whether the newly added sampling number of the disease diagnosis data reaches a preset data volume, and classifies the newly collected disease diagnosis data to obtain a medical training data set and a medical verification data set if the newly added sampling number exceeds the preset data volume. Note that the data amount of the disease diagnosis data in the medical training data set is larger than the data amount of the medical sampling data in the disease diagnosis data.
Furthermore, the medical training data set and the medical verification data set are divided to keep the consistency of the data distribution as much as possible. The medical training data set and the medical verification data set are divided according to the proportion of 7:3 as much as possible, so that the efficiency and the accuracy of subsequent clinical auxiliary model training are improved.
S300, updating the clinical assistant model according to the medical training data set and the medical verification data set, and updating the medical knowledge map according to the medical knowledge data to complete the learning and updating of the clinical assistant decision support system.
Specifically, after the CDSS system completes classification of the data set, the CDSS system updates the clinical auxiliary model according to the classified medical training data set and medical verification data set, and in addition, the CDSS system updates the medical knowledge map according to the medical knowledge data, so that the CDSS system completes learning and updating in time. Therefore, medical personnel in different regions and different regions can learn medical knowledge according to the timely and efficient updated CDSS system, the medical personnel can be assisted in diagnosing diseases, the accurate diagnosis of the diseases by the medical personnel is improved, and a treatment scheme with ideal curative effect on each patient is generated.
In addition, medical sampling data of different data sources are obtained through the expanded standardized interface, so that the medical sampling data in HIS of each hospital are mutually communicated, the later development cost is reduced, and the application and popularization of a CDSS system are facilitated. In addition, because the CDSS system obtains medical sampling data through the standardized interface, the user at the mobile terminal (e.g., mobile phone, watch, computer) can perform auxiliary diagnosis and decision-making based on the self-learning of the CDSS system without inputting a large amount of information by the user at the mobile terminal. The CDSS system can be used for self-learning to assist diagnosis and treatment, so that a user can clearly know disease information possibly suffered by the user, and can assist medical staff in diagnosing and mastering the disease information possibly suffered by the user.
In an embodiment of the present invention, as shown in fig. 2, a learning implementation method of a clinical assistant decision support system includes:
s110, acquiring disease diagnosis data and medical knowledge data through a standardized interface;
specifically, as shown in fig. 3, the CDSS system extends a plurality of standardized interfaces through which a plurality of data sources are connected, the data sources including, but not limited to, servers of medical institutions (e.g., general hospitals such as the hospital department, community health offices, and base-level health institutions such as village health rooms), servers of third-party institutions (e.g., health centers, health and welfare agencies), servers of medical information centers, and data servers corresponding to applications for developing medical services APP (e.g., doctor on-line APP, safe doctor APP, micro-doctor APP, etc.) to acquire disease diagnosis data and medical knowledge data.
The medical knowledge data includes, but is not limited to, clinical medical guidelines, medical literature data, and medical dictionaries. The disease diagnosis data includes, but is not limited to, the test result, Electronic Medical Record (EMR), and the content of any disease diagnosis data includes the basic information of the user. Electronic medical records are computerized medical record systems or computer-based patient records, and the content of the electronic medical records includes all information of paper medical records, such as admission records, medical records, video reports, and the like.
S120, structuring the disease diagnosis data and the medical knowledge data to accord with a preset data structure;
s130, generating a corresponding index identifier for each piece of disease diagnosis data, and counting the number of the index identifiers to obtain the number of newly added samples;
specifically, Natural Language Processing (NLP) is a sub-field of artificial intelligence AI, and disease diagnosis data and medical knowledge data are structured by Natural Language Processing technology to conform to a preset data structure.
After the CDSS system acquires the disease diagnosis data, the index identification corresponding to each disease diagnosis data is generated by adopting the corresponding index rule, and the CDSS system can obtain the newly added sampling number by counting the number of the index identifications because the index identification is unique and unrepeatable.
For example, after the disease diagnosis data is converted to conform to the preset data structure, the disease diagnosis data includes basic information of the user, and the basic information includes one or more of a user name, a user gender, a user identity card number, a user mobile phone number, a user medical insurance card number, a treatment card number, a treatment hospital, a treatment department, a treatment time, a medical examination time, and a hospital stay time. Accordingly, a corresponding unique index value can be generated from the basic information.
For example, after the disease diagnosis data is converted to conform to the preset data structure, a corresponding unique index value may be generated according to the acquisition timestamp of the disease diagnosis data, or a corresponding unique index value may be generated according to the conversion completion time of the disease diagnosis data converted into new medical sampling data conforming to the preset data structure. Of course, the corresponding unique index value may also be generated according to the timestamp and the at least two keywords with the highest frequency of occurrence of the disease diagnosis data.
The embodiment carries out standardized conversion on the original disease diagnosis data and medical knowledge data, namely, the disease diagnosis data and the medical knowledge data with different sources and different formats can be converted into standardized data by setting a uniform standard and a targeted conversion rule, so that the medical sampling data has better usability, and a favorable basis is provided for subsequent application.
Disease diagnostic data typically comes from multiple hospitals, medical facilities, or multiple departments of a hospital. The original disease diagnosis data is the original data generated in the user's treatment process, such as data of doctors recording user symptoms, data of results of medical examinations performed by users, and the like, and can be stored in HIS systems of various hospitals or documents such as paper medical records, so that the original disease diagnosis data can be extracted from the HIS database by interacting with HIS systems of various hospitals, and the paper data can be converted into the original disease diagnosis data in electronic format on the server by manual entry and the like.
Since the format of the generated original disease diagnosis data is different depending on HIS systems used in different hospitals or depending on the way and habit of recording data, etc., standardized conversion is required. During conversion, a uniform standard can be preset to convert the original disease diagnosis data into medical sampling data conforming to a preset data structure.
S200, when the newly added sampling number reaches a preset data volume, classifying the disease diagnosis data into a medical training data set and a medical verification data set;
s300, updating the clinical assistant model according to the medical training data set and the medical verification data set, and updating the medical knowledge map according to the medical knowledge data to complete the learning and updating of the clinical assistant decision support system.
Specifically, the CDSS system construction includes the following important processes: the first is to establish a medical knowledge base, and the data sources in the medical knowledge base comprise disease diagnosis data and medical knowledge data. Firstly, a medical knowledge base needs to be constructed by acquiring massive literature evidences and clinical practice evidences, and the medical knowledge base must be updated and maintained agilely along with the latest development of medicine. The second is to build a medical knowledge map based on the medical knowledge data. In general, deep learning by machine is used by AI to understand the medical logic, and like a doctor, to understand the whole process from suspicion to diagnosis to treatment of a disease, and how to make decision in the process according to the existing data and knowledge. And thirdly, establishing a clinical auxiliary model according to the disease diagnosis data.
After the CDSS system finishes the classification of the data set, the clinical auxiliary model is updated according to the classified medical training data set and the medical verification data set, so that the CDSS system can finish learning and updating in time. Therefore, medical staff in different areas and different regions can learn medical knowledge according to the CDSS system updated timely and efficiently. In addition, because the CDSS system can be embedded into the electronic medical records of the hospital in a web mode, when a doctor operates the electronic diseases, the CDSS system can provide various functions of medical knowledge base retrieval, treatment scheme recommendation, similar medical record recommendation, auxiliary diagnosis, medical advice quality control, clinical early warning and the like, and provides continuous support for the doctor from before diagnosis to during diagnosis to after diagnosis, so that the diagnosis efficiency and the diagnosis level of the doctor are improved, and the misdiagnosis rate is reduced.
In addition, medical sampling data of different data sources are obtained through the expanded standardized interface, so that the medical sampling data in HIS of each hospital are mutually communicated, the later development cost is reduced, and the application and popularization of a CDSS system are facilitated. In addition, because the CDSS system obtains medical sampling data through the standardized interface, the user at the mobile terminal (e.g., mobile phone, watch, computer) can perform auxiliary diagnosis and decision-making based on the self-learning of the CDSS system without inputting a large amount of information by the user at the mobile terminal. The CDSS system can be used for self-learning to assist diagnosis and treatment, so that a user can clearly know disease information possibly suffered by the user, and can assist medical staff in diagnosing and mastering the disease information possibly suffered by the user. The problem that the existing medical data management method depends on a single HIS and data intercommunication among different hospitals is difficult to realize is overcome at least to a certain extent through the embodiment.
In an embodiment of the present invention, as shown in fig. 4, a learning implementation method of a clinical assistant decision support system includes:
s110, acquiring disease diagnosis data and medical knowledge data through a standardized interface;
s121, judging whether the disease diagnosis data accord with medical knowledge or not according to the medical knowledge base;
s122, structuring the medical knowledge data and the disease diagnosis data conforming to the medical knowledge to conform to a preset data structure;
s123, data which accord with a preset data structure are subjected to duplicate checking, and duplicate data are deleted to obtain final disease diagnosis data and medical knowledge data;
specifically, it is described through the above example that the CDSS system creates the medical knowledge base, and once the CDSS system acquires the disease diagnosis data through the standardized interface, the acquired disease diagnosis data can be compared with the medical knowledge base, so as to determine whether the disease diagnosis data conforms to the medical knowledge. And if the disease diagnosis data does not conform to the medical knowledge, deleting and discarding the disease diagnosis data. If the disease diagnosis data conforms to the medical knowledge, the disease diagnosis data is retained. Then, the CDSS system performs the structuring process of the medical knowledge data and the disease diagnosis data conforming to the medical knowledge so as to conform to the preset data structure, in the manner of the above-described embodiment.
After the CDSS system acquires all candidate data which accord with a preset data structure, the data can be checked for duplication due to consistent data structures, so that the CDSS system can search repeated candidate data and delete the repeated candidate data, and new medical sampling data can be obtained.
S130, generating a corresponding index identifier for each piece of disease diagnosis data, and counting the number of the index identifiers to obtain the number of newly added samples;
s200, when the newly added sampling number reaches a preset data volume, classifying the disease diagnosis data into a medical training data set and a medical verification data set;
s310, inputting the medical training data set into an original clinical auxiliary model for training;
s320, adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
s330, updating medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data, and finishing the learning and updating of the clinical assistant decision support system;
specifically, the medical training data set is used for fitting and training a clinical auxiliary model, and when the newly added sampling number reaches a preset data volume, all the disease diagnosis data are classified into a medical training data set and a medical verification data set. Then, the CDSS system inputs the medical training data set to the original clinical assistant model, and trains the classified clinical assistant model by setting parameters of the clinical assistant model.
The medical verification data set is used for the CDSS system to obtain a plurality of candidate clinical auxiliary models through the training of the medical training data set, each candidate clinical auxiliary model is used for predicting data in the medical verification data set, each model is used for predicting the verification set data in order to find out the model with the best effect, the model accuracy is recorded, and the model parameter with the best effect is selected to be used for adjusting the model parameter of the original clinical auxiliary model to obtain the updated clinical auxiliary model.
Medical entities include, but are not limited to, diseases, symptoms, aliases, sites, departments, complications, medications, age, gender, exam-pies, expenses, and the like. The association relationship between the medical entity and the medical entity includes disease and symptom relationship, symptom and symptom relationship, disease and complication relationship, disease and alias relationship, disease and part relationship, disease and department relationship, disease and medicine relationship, disease and age relationship, disease and sex relationship, etc. And creating corresponding nodes according to the medical entities, creating node labels according to the names of the medical entities, connecting the nodes according to the incidence relation between the medical entities and the medical entities, and removing the duplication to complete the construction of the medical knowledge graph.
S400, inputting the obtained inquiry data into a clinical auxiliary model to obtain a probability value of each disease type;
s500, selecting a preset number of disease types with the probability values arranged in the front according to the probability values, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
Specifically, if a user passes through a medical service APP on a client (e.g., a mobile phone or a computer) and the medical service APP is connected with a CDSS system, the user can trigger the client to enter a dialogue inquiry when the inquiry request is received, and a dialogue inquiry box is displayed through a display interface of the client, so that the user inputs inquiry data of the user in the dialogue inquiry box, the inquiry data includes basic information and main complaint content, and the main complaint content is subjected to keyword extraction to obtain symptom information of the user. The basic information includes, but is not limited to, the age, sex, region, physical characteristics, allergen, past medical history, etc. of the user. The subject matter includes, but is not limited to, descriptive matter of the disease site, manifestation of the condition, and the like.
Since each disease may correspond to a symptom, and each symptom information may correspond to a plurality of types of disorders. After the client acquires the basic information and the symptom information through the dialogue inquiry box, the basic information and the symptom information of the user are transmitted to the CDSS system through the Internet, the CDSS system inputs the received basic information and the symptom information of the user to a clinical auxiliary model, the clinical auxiliary model can roughly deduce the disease information of the user, and the disease information comprises a disease type and a disease probability. The CDSS system selects disease types with the prior disease probability (for example, the first three), then the CDSS system inputs the disease types with the prior disease probability (for example, the first three) into the medical knowledge graph for comparison, so as to obtain the probability value of each disease type, then selects a preset number of disease types with the prior probability value arrangement according to the probability value of each auxiliary suggestion, outputs auxiliary suggestions (including but not limited to medication information and examination items) corresponding to the disease types and recommendation indexes of the auxiliary suggestions according to the medical knowledge graph and the disease types, and then the CDSS system selects a plurality of auxiliary suggestions with the prior recommendation indexes arrangement to generate final clinical auxiliary data.
Through the embodiment, the CDSS system can match with the disease symptom database according to the basic information and the symptom information to infer the disease information of the user, so that the user can clearly know the disease information possibly suffered by the user, and can assist a doctor to master the disease information possibly suffered by the user. Moreover, the CDSS system can assist the diagnosis and treatment decision of doctors and timely make clinical early warning, thereby improving the medical efficiency and relieving the diagnosis and treatment pressure of doctors. In addition, the CDSS system assists a doctor in making diagnosis and treatment decisions, so that the current situation that the professional level of the doctor is insufficient can be made up to a certain extent, the diagnosis and treatment level of the doctor is improved, the training progress is accelerated, and the doctor culture period is shortened. In addition, the CDSS can help the primary level to establish a homogenization and standardized medical path, help primary level doctors to avoid misdiagnosis and missed diagnosis of common diseases, and help primary level doctors to carry out scientific referral, thereby being beneficial to improving the quality of primary level medical services, better promoting implementation of a grading diagnosis and treatment policy and relieving the current situation of uneven medical resource allocation.
According to the invention, through collecting data and enabling the CDSS system to learn and update in real time, credible, timely updated and digitalized diagnosis and treatment knowledge can be provided for medical care personnel for common diseases at any time and any place, and the medical care personnel can be helped, so that the medical level of primary doctors can be greatly improved, and the assistance is improved for national promoted graded diagnosis and treatment, thereby greatly reducing the cost of medical service, ensuring the fairness and accessibility of the medical service, and possibly enabling people to enjoy high-quality medical service. The CDSS system presents medical knowledge that the medical personnel can not remember through a computer interface to help the medical personnel to see a doctor. The updating frequency and the convenience degree of the CDSS system are far from being compared with paper data, and the scientificity, the safety, the effectiveness and the applicability of disease diagnosis and treatment can be enhanced by updating the clinical auxiliary model in the CDSS system in time, so that the realization of individualized medical treatment is promoted, the reliability of the CDSS system is improved, and the popularization of the CDSS system is facilitated. The CDSS system can also prompt incompatibility and anaphylactic reaction of medicines, so that doctors can learn a medicine administration scheme and ensure the safety of medicine administration, the diagnosis capability of the doctors can be greatly improved, the misdiagnosis rate is greatly reduced, and good medical service is provided for users. The invention can utilize CDSS to construct county-area integrated medical systems, integrate multi-party resources and coordinate benefits of all parties, so as to construct a benign regional medical ecological system by taking the improvement of the diagnosis and treatment capability of doctors in the integrated medical systems and the service capability of medical institutions as guidance.
For example, a conventional clinical decision assistant system (CDSS) is typically deeply tied to a hospital information management system (HIS) or is a separate mobile phone application. The upgrading of deep binding is difficult, HISs of hospitals are not communicated with each other, the later development cost is extremely high, and the application and popularization of the CDSS are not facilitated. Independent mobile phone software can carry out auxiliary diagnosis and decision-making only after inputting a large amount of information of a patient.
The hospital-based HIS or EMR plug-in system does not need to be deeply bound with the original system. The hospital system is only required to provide a standardized interface (such as an API), the inquiry data is input into the cloud or a localized CDSS system through the API, the calculation result can be automatically given, and due to the fact that the HIS system or the EMR system is accessed through the API, data sharing across the HIS system or the EMR system can be achieved, and the data interaction barrier is broken.
The CDSS system can be based on the input inquiry data, and has a mode of supervised learning, namely, the CDSS system stores data by itself, which is called a modeling queue (corresponding to the original clinical auxiliary model of the present invention), and new input data (corresponding to the new medical sampling data of the present invention), which is called an upgrade queue (corresponding to the medical training data set of the present invention) and a verification queue (corresponding to the medical verification data set of the present invention). When the data volume of the external input approaches the modeling queue (namely, the data volume is equivalent to the situation that the newly added sampling number reaches the preset data volume), the model is automatically relearned and modeled, and a new clinical auxiliary model is generated. By repeating the steps, multi-dimensional comprehensive judgment can be provided according to the clinical auxiliary model and the medical knowledge map which are updated in time, and the method is not only used for single disease or single disease species. The CDSS system can further verify the sensitivity and specificity of the CDSS system according to the input data, and meanwhile, conditions are provided for the optimization of the CDSS.
In addition, the CDSS system stores a large amount of desensitization data, and through inputting inquiry data externally, supervised learning is carried out by combining the original system of the CDSS, and self optimization is continuously carried out, so that the reliability and the accuracy of the auxiliary decision of the CDSS system are improved.
In one embodiment of the present invention, as shown in fig. 5, a learning implementation system of a clinical assistant decision support system includes:
the acquisition module 10 is used for acquiring new disease diagnosis data and medical knowledge data through a standardized interface and counting the newly added sampling number of the disease diagnosis data;
an obtaining module 20, configured to classify the disease diagnosis data into a medical training data set and a medical verification data set when the newly added sampling number reaches a preset data amount;
and the learning module 30 is configured to update the clinical assistant model according to the medical training data set and the medical verification data set, and update the medical knowledge map according to the medical knowledge data to complete learning and updating of the clinical assistant decision support system.
Specifically, this embodiment is a system embodiment corresponding to the above method embodiment, and specific effects refer to the above method embodiment, which are not described in detail herein.
Based on the foregoing embodiment, the acquisition module 10 includes:
the acquisition unit is used for acquiring disease diagnosis data and medical knowledge data through a standardized interface;
the data processing unit is used for carrying out structuring processing on the disease diagnosis data and the medical knowledge data so as to accord with a preset data structure;
and the generation counting unit is used for generating corresponding index identification for each piece of disease diagnosis data and counting the number of the index identifications to obtain the newly added sampling number.
Specifically, this embodiment is a system embodiment corresponding to the above method embodiment, and specific effects refer to the above method embodiment, which are not described in detail herein.
Based on the foregoing embodiment, the data processing unit includes:
the judging subunit is used for judging whether the disease diagnosis data accords with the medical knowledge according to the medical knowledge base;
the processing subunit is used for carrying out structural processing on the medical knowledge data and the disease diagnosis data conforming to the medical knowledge so as to conform to a preset data structure;
and the screening and removing subunit is used for carrying out duplicate checking on the data which accord with the preset data structure and deleting the duplicate data to obtain final disease diagnosis data and medical knowledge data.
Specifically, this embodiment is a system embodiment corresponding to the above method embodiment, and specific effects refer to the above method embodiment, which are not described in detail herein.
Based on the foregoing embodiment, the learning module 30 includes:
the training unit is used for inputting the medical training data set into an original clinical auxiliary model for training;
the verification updating unit is used for adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
and the map updating unit is used for updating the medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data and finishing the learning and updating of the clinical assistant decision support system.
Specifically, this embodiment is a system embodiment corresponding to the above method embodiment, and specific effects refer to the above method embodiment, which are not described in detail herein.
Based on the foregoing embodiment, further comprising:
the input module is used for acquiring inquiry data;
the processing module is used for inputting the acquired inquiry data into the clinical auxiliary model and acquiring the probability value of each disease type; and selecting a preset number of disease types with the probability values arranged in the front according to the probability values, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
Specifically, this embodiment is a system embodiment corresponding to the above method embodiment, and specific effects refer to the above method embodiment, which are not described in detail herein.
It will be apparent to those skilled in the art that for convenience and brevity of description, only the above-described division of program modules is illustrated, and in actual practice, the above-described distribution of functions may be performed by different program modules, that is, the internal structure of the apparatus may be divided into different program units or modules to perform all or part of the above-described functions. Each program module in the embodiments may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one processing unit, and the integrated units may be implemented in a form of hardware, or may be implemented in a form of software program unit. In addition, the specific names of the program modules are only used for distinguishing one program module from another, and are not used for limiting the protection scope of the application.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to the related descriptions of other embodiments for parts that are not described or recited in detail in a certain embodiment.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus/client and method may be implemented in other ways. For example, the above-described apparatus/client embodiments are merely illustrative, and for example, the division of the modules or units is only one logical division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may also be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
It should be noted that the above embodiments can be freely combined as necessary. The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (10)
1. A learning implementation method of a clinical assistant decision support system is characterized by comprising the following steps:
acquiring new disease diagnosis data and medical knowledge data through a standardized interface, and counting the newly added sampling number of the disease diagnosis data;
when the newly added sampling number reaches a preset data volume, classifying the disease diagnosis data into a medical training data set and a medical verification data set;
and updating a clinical assistant model according to the medical training data set and the medical verification data set, and updating a medical knowledge map according to the medical knowledge data to finish the learning and updating of the clinical assistant decision support system.
2. The learning realization method of clinical assistant decision support system according to claim 1, wherein the step of obtaining new disease diagnosis data and medical knowledge data through standardized interface and counting the number of newly added samples of the disease diagnosis data comprises the steps of:
acquiring disease diagnosis data and medical knowledge data through a standardized interface;
structuring the disease diagnosis data and the medical knowledge data to conform to a preset data structure;
and generating corresponding index identification for each piece of disease diagnosis data, and counting the number of the index identifications to obtain the newly added sampling number of the medical sampling data.
3. The learning realization method of clinical assistant decision support system according to claim 2, wherein the structuring of the disease diagnosis data and medical knowledge data to fit a preset data structure comprises the steps of:
judging whether the disease diagnosis data conforms to medical knowledge according to a medical knowledge base;
structuring the medical knowledge data and the disease diagnosis data conforming to medical knowledge to conform to a preset data structure;
and (4) checking the data which accord with the preset data structure for duplication and deleting the duplicated data to obtain the final disease diagnosis data and medical knowledge data.
4. The learning realization method of clinical assistant decision support system according to claim 1, wherein the updating of the clinical assistant model based on the medical training data set and the medical validation data set and the updating of the medical knowledge map based on the medical knowledge data to complete the learning update of the clinical assistant decision support system comprises the steps of:
inputting the medical training data set into an original clinical auxiliary model for training;
adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
and updating the medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data, and finishing the learning and updating of the clinical assistant decision support system.
5. The learning realization method of a clinical assistant decision support system according to any of the claims 1-4, characterized by further comprising the steps of:
inputting the obtained inquiry data into the clinical auxiliary model, and obtaining the probability value of each disease type;
and selecting a preset number of disease types with the probability values arranged in the front according to the probability values, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
6. A learning implementation system of a clinical assistant decision support system, comprising:
the acquisition module is used for acquiring new disease diagnosis data and medical knowledge data through the standardized interface and counting the newly added sampling number of the disease diagnosis data;
the acquisition module is used for classifying the disease diagnosis data into a medical training data set and a medical verification data set when the newly added sampling number reaches a preset data volume;
and the learning module is used for updating the clinical auxiliary model according to the medical training data set and the medical verification data set and updating the medical knowledge map according to the medical knowledge data so as to complete the learning and updating of the clinical auxiliary decision support system.
7. The learning implementation system of a clinical assistant decision support system according to claim 6, wherein the acquisition module comprises:
the acquisition unit is used for acquiring disease diagnosis data and medical knowledge data through a standardized interface;
the data processing unit is used for carrying out structural processing on the disease diagnosis data and the medical knowledge data so as to accord with a preset data structure;
and the generation counting unit is used for generating corresponding index identification for each piece of disease diagnosis data, and counting the number of the index identifications to obtain the newly added sampling number of the medical sampling data.
8. The learning realization system of a clinical assistant decision support system according to claim 7, characterized in that the data processing unit comprises:
the judging subunit is used for judging whether the disease diagnosis data conforms to medical knowledge or not according to a medical knowledge base;
the processing subunit is used for carrying out structural processing on the medical knowledge data and the disease diagnosis data conforming to the medical knowledge so as to conform to a preset data structure;
and the screening and removing subunit is used for carrying out duplicate checking on the data which accord with the preset data structure and deleting the repeated data to obtain the final disease diagnosis data and medical knowledge data.
9. The learning implementation system of the clinical assistant decision support system according to claim 6, wherein the learning module comprises:
the training unit is used for inputting the medical training data set into an original clinical auxiliary model for training;
the verification updating unit is used for adjusting model parameters corresponding to the trained clinical auxiliary model according to the medical verification data set to obtain an updated clinical auxiliary model;
and the map updating unit is used for updating the medical entities in the medical knowledge map and the incidence relation of each medical entity according to the medical knowledge data and finishing the learning and updating of the clinical assistant decision support system.
10. The learning realization system of a clinical assistant decision support system according to any of the claims 6-9, further comprising:
the input module is used for acquiring inquiry data;
the processing module is used for inputting the acquired inquiry data into the clinical auxiliary model and acquiring the probability value of each disease type; and selecting a preset number of disease types with the probability values arranged in the front according to the probability values, and outputting final clinical auxiliary data according to the medical knowledge map and the disease types.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011231368.0A CN112365974A (en) | 2020-11-06 | 2020-11-06 | Learning implementation method and system of clinical assistant decision support system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011231368.0A CN112365974A (en) | 2020-11-06 | 2020-11-06 | Learning implementation method and system of clinical assistant decision support system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112365974A true CN112365974A (en) | 2021-02-12 |
Family
ID=74508962
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011231368.0A Pending CN112365974A (en) | 2020-11-06 | 2020-11-06 | Learning implementation method and system of clinical assistant decision support system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112365974A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112951414A (en) * | 2021-03-26 | 2021-06-11 | 嘉兴市第二医院 | Primary medical clinical assistant decision-making system |
| CN113555110A (en) * | 2021-07-15 | 2021-10-26 | 北京鹰瞳科技发展股份有限公司 | Method and equipment for training multi-disease referral model |
| CN113628734A (en) * | 2021-08-02 | 2021-11-09 | 浙江海心智惠科技有限公司 | Design method of oncology electronic medical advice system with clinical decision intelligent recommendation function |
| CN114496143A (en) * | 2021-12-31 | 2022-05-13 | 医渡云（北京）技术有限公司 | Method and device for associating clinical data with literature, electronic equipment and storage medium |
| CN117198505A (en) * | 2023-08-23 | 2023-12-08 | 深圳大学 | Deep learning language model fine tuning method for clinical medicine decision assistance |
| CN117238455A (en) * | 2023-09-18 | 2023-12-15 | 安徽亚创电子科技有限责任公司 | Hospital diagnosis supervision optimization system based on Internet of things |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108427735A (en) * | 2018-02-28 | 2018-08-21 | 东华大学 | Clinical knowledge map construction method based on electronic health record |
| CN109346169A (en) * | 2018-10-17 | 2019-02-15 | 长沙瀚云信息科技有限公司 | A kind of artificial intelligence assisting in diagnosis and treatment system and its construction method, equipment and storage medium |
| CN111816304A (en) * | 2020-07-22 | 2020-10-23 | 北京聚道科技有限公司 | Method and system for establishing genetic disease assistant decision |
-
2020
- 2020-11-06 CN CN202011231368.0A patent/CN112365974A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108427735A (en) * | 2018-02-28 | 2018-08-21 | 东华大学 | Clinical knowledge map construction method based on electronic health record |
| CN109346169A (en) * | 2018-10-17 | 2019-02-15 | 长沙瀚云信息科技有限公司 | A kind of artificial intelligence assisting in diagnosis and treatment system and its construction method, equipment and storage medium |
| CN111816304A (en) * | 2020-07-22 | 2020-10-23 | 北京聚道科技有限公司 | Method and system for establishing genetic disease assistant decision |
Non-Patent Citations (2)
| Title |
|---|
| 樊小玲主编: "《军队卫生信息资源开发利用研究与实践》", 30 June 2012, 人民军医出版社 * |
| 黄玲: "《JavaEE程序设计及项目开发教程 JSP篇》", 31 January 2017, 重庆大学出版社 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112951414A (en) * | 2021-03-26 | 2021-06-11 | 嘉兴市第二医院 | Primary medical clinical assistant decision-making system |
| CN113555110A (en) * | 2021-07-15 | 2021-10-26 | 北京鹰瞳科技发展股份有限公司 | Method and equipment for training multi-disease referral model |
| CN113628734A (en) * | 2021-08-02 | 2021-11-09 | 浙江海心智惠科技有限公司 | Design method of oncology electronic medical advice system with clinical decision intelligent recommendation function |
| CN114496143A (en) * | 2021-12-31 | 2022-05-13 | 医渡云（北京）技术有限公司 | Method and device for associating clinical data with literature, electronic equipment and storage medium |
| CN117198505A (en) * | 2023-08-23 | 2023-12-08 | 深圳大学 | Deep learning language model fine tuning method for clinical medicine decision assistance |
| CN117238455A (en) * | 2023-09-18 | 2023-12-15 | 安徽亚创电子科技有限责任公司 | Hospital diagnosis supervision optimization system based on Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112365974A (en) | Learning implementation method and system of clinical assistant decision support system | |
| CN110249392A (en) | Intelligent assisting in diagnosis and treatment system and method | |
| CN111599488A (en) | Intelligent inquiry implementing method, system and storage medium | |
| US20040044546A1 (en) | Checklist-based flow and tracking system for patient care by medical providers | |
| US20120185275A1 (en) | System and method of automated data analysis for implementing health records personal assistant with automated correlation of medical services to insurance and tax benefits for improved personal health cost management | |
| WO2021159761A1 (en) | Pathological data analysis method and apparatus, and computer device and storage medium | |
| CN112133445A (en) | Cardiovascular disease management service method and system | |
| WO2022267678A1 (en) | Video consultation method and apparatus, device and storage medium | |
| CN106650256A (en) | Precise medical platform for molecular diagnosis and treatment | |
| CN103339605A (en) | Managing healthcare information in a distributed system | |
| Chen et al. | Inferring clinical workflow efficiency via electronic medical record utilization | |
| CN114187988A (en) | Data processing method, device, system and storage medium | |
| CN113345583A (en) | Method and system for constructing full life cycle resident intelligent health file | |
| JP6908977B2 (en) | Medical information processing system, medical information processing device and medical information processing method | |
| EP3506315A1 (en) | Method of using medical data related to patients suffering a given disease | |
| CN112117002A (en) | Novel intelligent psychological assessment intervention system and method combined with virtual reality technology | |
| Pavlov et al. | Reference medical datasets (MosMedData) for independent external evaluation of algorithms based on artificial intelligence in diagnostics | |
| CN114121266A (en) | Intelligent auxiliary diagnosis method and system | |
| CN109192312B (en) | Intelligent management system and method for adverse events of heart failure patients | |
| KR19980025157A (en) | Disease Management Methods and Systems | |
| Dillard et al. | Insights into conducting audiological research with clinical databases | |
| Wikansari et al. | What are the barriers to the implementation of electronic medical records? A review of recent studies | |
| KR20180002229A (en) | An agent apparatus for constructing database for dementia information and the operating method by using the same | |
| CN102521501A (en) | Simulation doctor system of intelligent medicine based on composite diagnosis and treatment template | |
| Li et al. | Predicting the loss to follow‐up (LTFU) of HIV/AIDS patients in China using a recency‐frequency (RF) model |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210212 |