CN117059216A - Medical clinical result prediction system generation device, method, equipment and medium - Google Patents
Medical clinical result prediction system generation device, method, equipment and medium Download PDFInfo
- Publication number
- CN117059216A CN117059216A CN202311045311.5A CN202311045311A CN117059216A CN 117059216 A CN117059216 A CN 117059216A CN 202311045311 A CN202311045311 A CN 202311045311A CN 117059216 A CN117059216 A CN 117059216A
- Authority
- CN
- China
- Prior art keywords
- medical
- data
- training
- structured
- training set
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000012549 training Methods 0.000 claims abstract description 255
- 230000004927 fusion Effects 0.000 claims description 24
- 238000002372 labelling Methods 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 14
- 238000007781 pre-processing Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 11
- 238000013075 data extraction Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000011282 treatment Methods 0.000 description 69
- 201000010099 disease Diseases 0.000 description 28
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 28
- 238000003745 diagnosis Methods 0.000 description 18
- 206010067484 Adverse reaction Diseases 0.000 description 17
- 230000006838 adverse reaction Effects 0.000 description 17
- 238000011160 research Methods 0.000 description 16
- 230000000694 effects Effects 0.000 description 13
- 230000008569 process Effects 0.000 description 13
- 239000003814 drug Substances 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 238000007418 data mining Methods 0.000 description 7
- 238000011269 treatment regimen Methods 0.000 description 6
- 230000036541 health Effects 0.000 description 5
- 229940079593 drug Drugs 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 208000028659 discharge Diseases 0.000 description 3
- 230000010354 integration Effects 0.000 description 3
- 238000012216 screening Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000006757 chemical reactions by type Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000890 drug combination Substances 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000002105 nanoparticle Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000007170 pathology Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000001959 radiotherapy Methods 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
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
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
-
- 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
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The application discloses a medical clinical result prediction system generating device, a method, equipment and a medium, relating to the technical field of computers, comprising the following steps: the medical text database creation module is used for extracting data of the acquired medical data and creating a medical text database; the training set generation module is used for generating a medical paper structured training set and a medical document structured training set; the structural database creation module is used for training the model through the medical paper structural training set and the medical document structural training set to generate a medical paper structural database and a medical document structural database; the system generation module is used for integrating the generated structured database and generating a clinical result prediction system through the obtained fused structured database. Therefore, the clinical result can be predicted by combining the constructed clinical result prediction system with the medical paper data and the actual clinical data, and the accuracy of the clinical result prediction is improved.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a device, a method, an apparatus, and a medium for generating a medical clinical result prediction system.
Background
With the rapid development of current medical technology and clinical research, the understanding and classification of diseases is also in progress, leading to a more and more differentiation of treatment schemes for specific diseases and disease subtypes, and treatment drug types in treatment schemes, and/or drug combination modes, and/or combination modes of drugs and other treatment means such as surgery, etc. At the same time, as new treatment regimens are developed rapidly, patient inclusion criteria in a study of the efficacy of a particular treatment regimen are finer. Therefore, in actual clinical work, a method and a system are needed for assisting doctors in providing detailed disease diagnosis suggestions for medical workers according to analysis of physiological characteristics of patients and predicting specific curative effects and adverse reactions of various existing treatment schemes for diagnosing the detailed diseases, so that the continuously-increased disease treatment schemes are reasonably selected and applied.
In the prior art, the prediction of clinical results is mostly realized by the fusion of technologies such as clinical medicine and computer artificial intelligence, but the method adopted in the prior art cannot be based on the historical data such as real world medical documents, comprehensive treatment schemes, basic characteristics of patients, adverse reactions in the process, treatment effects and the like, data mining is carried out on clinical results such as adverse reactions and treatment effects after the diagnosis of refined diseases and dynamic adjustment of the treatment schemes of diseases, and the data mining of clinical documents only aims at exploring medicine indications, and the mining of the correlation of the treatment schemes, the treatment effects, the adverse reactions and the like in the clinical document data is not yet available.
Disclosure of Invention
In view of the above, the present application aims to provide a device, a method, a device and a medium for generating a medical clinical result prediction system, which can predict a clinical result by combining medical paper data and actual clinical data through a constructed clinical result prediction system, and improve the accuracy of clinical result prediction. The specific scheme is as follows:
in a first aspect, the present application provides a medical clinical outcome prediction system generation apparatus, comprising:
the medical text database creation module is used for carrying out data extraction on the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data;
the training set generation module is used for generating a medical paper structured training set based on the preprocessed medical paper data and generating a medical document structured training set based on the preprocessed medical document data;
the structured database creation module is used for training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively so as to generate a corresponding medical paper structured database and a corresponding medical document structured database;
And the system generation module is used for integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system.
Optionally, the medical text database creation module includes:
the text data preprocessing unit is used for acquiring medical data, extracting medical paper data and medical document data in the medical data, and respectively marking attributes of the medical paper data and the medical document data to obtain preprocessed medical paper data and preprocessed medical document data;
and the database creation unit is used for creating a medical text database based on the preprocessed medical paper data and the preprocessed medical document data.
Optionally, the training set generating module includes:
the data attribute determining unit is used for acquiring the preprocessed medical paper data and the preprocessed medical document data stored in the medical text database and respectively determining a relation between a first entity type and a first entity in the preprocessed medical paper data and a relation between a second entity type and a second entity in the medical document data;
The first training data determining unit is used for creating a first data tag based on the relation between the first entity type and the first entity, marking the preprocessed medical paper data through the first data tag, and taking the marked medical paper data as first training data;
the second training data determining unit is used for creating a second data tag based on the relation between the second entity type and the second entity, marking the preprocessed medical document data through the second data tag, and taking the marked medical document data as second training data;
and the training set generating unit is used for respectively carrying out standardization operation and normalization operation on the first training data and the second training data so as to respectively generate a medical paper structured training set and a medical document structured training set.
Optionally, the medical clinical result prediction system generating device further includes:
the model training module is used for training a preset model to be trained based on the historical training set so as to obtain a preset training corpus model.
Optionally, the structured database creation module includes:
A first model training set generating unit, configured to generate a first pre-training set and a first fine tuning training set based on the medical paper structured training set;
the first model pre-training unit is used for training the preset training corpus model through the first pre-training set so as to obtain a first pre-trained model;
the first model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the first pre-trained model through the first fine tuning training set so as to obtain a first labeling model;
and the medical paper structured database generation unit is used for processing the medical paper data in the medical text database through the first labeling model so as to obtain the medical paper structured database.
Optionally, the structured database creation module includes:
a second model training set generating unit, configured to generate a second pre-training set and a second fine tuning training set based on the medical document structured training set;
the second model pre-training unit is used for training the preset training corpus model through the second pre-training set so as to obtain a second pre-trained model;
The second model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the second pre-trained model through the second fine tuning training set so as to obtain a second labeling model;
and the medical document structured database generation unit is used for extracting the time tag of the attribute data in the medical document structured training set, and processing the medical document data in the medical text database based on the time tag and the second annotation model to obtain the medical document structured database.
Optionally, the system generating module includes:
the first entity relation set generation unit is used for matching preset medical document data based on the medical paper structured database and generating a first entity relation set through the matched entity type and entity relation;
the second entity relationship set generating unit is used for matching preset medical document data based on the medical document structured database and generating a second entity relationship set through the matched entity type and entity relationship;
a third entity relationship set generating unit, configured to perform entity type filtering and entity relationship correction on the second entity relationship set based on the first entity relationship set, so as to obtain a target entity relationship set;
And the system generation unit is used for establishing a fusion structured database based on the target entity relation set, and generating a clinical result prediction system based on the fusion structured database and a preset retrieval tool so as to predict clinical results through the clinical result prediction system.
In a second aspect, the present application provides a medical clinical outcome prediction system generation method, comprising:
extracting data from the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data;
generating a medical paper structured training set based on the preprocessed medical paper data, and generating a medical document structured training set based on the preprocessed medical document data;
training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively to generate a corresponding medical paper structured database and a corresponding medical document structured database;
and integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system.
In a third aspect, the present application provides an electronic device, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the aforementioned medical clinical outcome prediction system generation method.
In a fourth aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the medical clinical outcome prediction system generation method described above.
In the application, a medical text database creation module is used for extracting data of acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data; the training set generation module is used for generating a medical paper structured training set based on the preprocessed medical paper data and generating a medical document structured training set based on the preprocessed medical document data; the structured database creation module is used for training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively so as to generate a corresponding medical paper structured database and a corresponding medical document structured database; and the system generation module is used for integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system. Therefore, the method and the system respectively preprocess the acquired medical paper data and the medical document data, generate corresponding structured training sets by utilizing the preprocessed data, train the model by utilizing the structured training sets, generate a medical paper structured database and a medical document structured database, and finally integrate the medical paper structured database and the medical document structured database into a database and construct a clinical result prediction system. Thus, after the literature data in the medical paper data are extracted, the clinical result can be predicted by combining the authoritative literature data and the actual clinical data, and the accuracy of the clinical result prediction can be effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present application, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a medical clinical outcome prediction system generating device provided by the application;
FIG. 2 is a timing diagram generated by the medical clinical outcome prediction system provided by the application;
FIG. 3 is a timing diagram of generating an annotation model according to the present application;
FIG. 4 is a schematic diagram of fine tuning of an annotation model according to the present application;
FIG. 5 is a schematic diagram of a specific medical clinical outcome prediction system generating device according to the present application;
FIG. 6 is a flowchart of a method for generating a medical clinical outcome prediction system provided by the present application;
fig. 7 is a block diagram of an electronic device according to the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
In the prior art, the prediction of clinical results is mostly realized by the fusion of technologies such as clinical medicine and computer artificial intelligence, but the method adopted in the prior art cannot be based on the historical data such as real world medical documents, comprehensive treatment schemes, basic characteristics of patients, adverse reactions in the process, treatment effects and the like, data mining is carried out on clinical results such as adverse reactions and treatment effects after the diagnosis of refined diseases and dynamic adjustment of the treatment schemes of diseases, and the data mining of clinical documents only aims at exploring medicine indications, and the mining of the correlation of the treatment schemes, the treatment effects, the adverse reactions and the like in the clinical document data is not yet available.
In order to solve the technical problems, the application provides a device, a method, equipment and a medium for generating a medical clinical result prediction system, which can predict a clinical result by combining medical paper data and actual clinical data through a constructed clinical result prediction system, and improve the accuracy of clinical result prediction.
Referring to fig. 1, an embodiment of the present application discloses a medical clinical outcome prediction system generating device, including:
the medical text database creation module 11 is configured to perform data extraction on acquired medical data, perform preprocessing on the acquired medical paper data and medical document data, and create a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data.
In this embodiment, a medical text database creation module, where the medical text database creation module includes: the text data preprocessing unit is used for acquiring medical data, extracting medical paper data and medical document data in the medical data, and respectively marking attributes of the medical paper data and the medical document data to obtain preprocessed medical paper data and preprocessed medical document data; and the database creation unit is used for creating a medical text database based on the preprocessed medical paper data and the preprocessed medical document data. That is, as shown in fig. 2, if a clinical result prediction system is to be constructed, medical data needs to be acquired first, and it is to be noted that the medical data is medical paper data in an electronic document format, medical records data in an electronic medical record system of medical staff, and image data obtained by photographing or capturing a daily clinical medical document of a medical paper or medical staff, after the medical data is acquired, text data and image data in the medical data need to be extracted, and the extracted text data includes the medical paper data and the medical document data, and after the text data is extracted, the extracted text data needs to be preprocessed, and various attribute tags can be created as required, so as to perform adjustment, editing, new addition, classification, sorting, retrieval, screening and returning processing in the data processing, storage and use processes. And after the attribute tag creation is completed, a medical text database may be created based on the preprocessed medical paper data and the preprocessed medical document data. For example, medical document data includes, but is not limited to, medical records, examinations, tests, communications, health records, medical records, health tests, or health monitors, and medical paper data includes, but is not limited to, research reports, research records, literature, guidelines, books, journals, and the like. Further, attribute tags created from medical document data include, but are not limited to, medical records first page, patient basic information, time, complaints, current medical history, past history, auxiliary exam, primary diagnosis, admission diagnosis, specialty, diagnosis and treatment history, discharge symptoms or signs, medical orders, exam report, pathology report, medical follow-up record; attribute tags created for medical paper data include, but are not limited to, abstracts, research backgrounds, nano-sized populations, research protocols, intervention/treatment protocols, adverse reactions, treatment effects, research discussions, research records, related tabular data. After successful creation of the attribute tags for the medical paper data and the medical document data, i.e. after the preprocessing is completed, a medical text database may be created based on the preprocessed medical paper data and the preprocessed medical document data.
It should be further noted that the medical data may further include medical picture data, and when the medical data is processed, the medical picture data in the medical data may be identified and extracted, and a medical picture database may be created based on the medical picture data that is identified and extracted.
The training set generating module 12 is configured to generate a medical paper structured training set based on the preprocessed medical paper data, and generate a medical document structured training set based on the preprocessed medical document data.
In this embodiment, the training set generating module includes: the data attribute determining unit is used for acquiring the preprocessed medical paper data and the preprocessed medical document data stored in the medical text database and respectively determining a relation between a first entity type and a first entity in the preprocessed medical paper data and a relation between a second entity type and a second entity in the medical document data; the first training data determining unit is used for creating a first data tag based on the relation between the first entity type and the first entity, marking the preprocessed medical paper data through the first data tag, and taking the marked medical paper data as first training data; the second training data determining unit is used for creating a second data tag based on the relation between the second entity type and the second entity, marking the preprocessed medical document data through the second data tag, and taking the marked medical document data as second training data; and the training set generating unit is used for respectively carrying out standardization operation and normalization operation on the first training data and the second training data so as to respectively generate a medical paper structured training set and a medical document structured training set. That is, after the medical text database is created, the medical paper structured training set and the medical document structured training set can be created respectively by using the preprocessed medical paper data and the preprocessed medical document data stored in the medical text database. The medical paper structured training set and the medical document structured training set are created, firstly, part of the preprocessed medical paper data and part of the preprocessed medical document data need to be extracted from a medical text database, and the extracted part of the preprocessed medical paper data and part of the preprocessed medical document data need to be subjected to clinical data mining and clinical outcome prediction, attribute labels corresponding to the clinical data mining and clinical outcome prediction purposes are determined, and the preprocessed medical paper data and the preprocessed medical document data containing the attribute labels are extracted as data for constructing the structured training set. After the data extraction is completed, a first entity type and a first entity relationship in the extracted part of the preprocessed medical paper data are required to be determined respectively, a second entity type and a second entity relationship in the extracted part of the preprocessed medical document data are determined, then a first data tag is created based on the determined first entity type and the first entity relationship, and a second data tag is created based on the determined second entity type and the second entity relationship. After the data tag is created, the extracted part of the preprocessed medical paper data and the preprocessed medical document data are marked according to the created first data tag and the second data tag respectively, so that first training data corresponding to the medical paper data and second training data corresponding to the medical document data are obtained.
For example, in the process of constructing a structured training set of medical papers, entity type labels are required to be created by using the names of treatment schemes, the periods and the frequencies of the treatment schemes, the treatment medicaments, operation means such as operations, interventions and radiotherapy and the like contained in the treatment schemes, the specific doses of the treatment medicaments and the means, the treatment time and other entity types in the process of acquiring content attribute data of the treatment schemes and the related tables, the entity relationship labels are created by using the clinical relationship among the entity types as entity relationships, and training data are marked by the created entity type labels and the entity relationship labels; in the same scheme, in the research background, the content attribute part of the inclusion crowd and the related form data, the basic physiological characteristics, the disease progress characteristics, the previous disease treatment history and the like of the research crowd are marked with entity types and entity relations; in the adverse reaction and the content attribute part of the related form data, labeling entity types and entity relationships of the types, grades, the number of people and the like of the adverse reaction after treatment; and marking entity types and entity relations by using the treatment effect evaluation end point type, end point result data statistics data and the like in the total of the treatment effect and the content attribute part of the related table. After labeling, the labeled data needs to be standardized and normalized to obtain the structured training set of the medical paper.
For example, a structured training set of medical documents is constructed, in the acquired content attribute data such as a course record, an in-patient doctor's advice and the like, the name of a treatment scheme related to clinical practice, the implementation times and specific time of the treatment scheme are used, the treatment drugs, operation, intervention and other operation means contained in each treatment scheme, the specific doses and treatment time of the treatment drugs and means are used as entity type labels, the clinical relationship among the entities is used as entity relationship labels, and training data are marked through the created entity type labels and entity relationship labels; the same scheme is used for marking entity types and entity relations of basic physiological characteristics of patients in content attribute parts such as basic information of the patients, auxiliary examination, examination reports with labels as hospital attributes and the like; in the content attribute parts of the current medical history, the past history and the like, marking entity types and entity relations for the treatment history and the like of the patient diseases; in the content attribute parts of main complaints, preliminary diagnosis, admission diagnosis, special cases, pathological reports and the like, marking entity types and entity relations of patient disease characteristics and the like; marking entity types and entity relations of adverse reaction types, adverse reaction levels, adverse reaction occurrence time, treatment modes of adverse reactions each time, treatment end point types, treatment end point time reaching each treatment end point time and the like of patients in content attribute parts such as discharge symptoms or physical signs, discharge orders, medical follow-up records, inspection reports with labels being out-of-hospital attributes and the like; after the labeling is completed, the labeled data needs to be standardized and normalized to obtain the medical document structured training set.
The structured database creation module 13 is configured to train a preset training corpus model through the medical paper structured training set and the medical document structured training set, so as to generate a corresponding medical paper structured database and a corresponding medical document structured database.
In this embodiment, the structured database creation module includes: a first model training set generating unit, configured to generate a first pre-training set and a first fine tuning training set based on the medical paper structured training set; the first model pre-training unit is used for training the preset training corpus model through the first pre-training set so as to obtain a first pre-trained model; the first model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the first pre-trained model through the first fine tuning training set so as to obtain a first labeling model; and the medical paper structured database generation unit is used for processing the medical paper data in the medical text database through the first labeling model so as to obtain the medical paper structured database.
In this embodiment, the structured database creation module includes: a second model training set generating unit, configured to generate a second pre-training set and a second fine tuning training set based on the medical document structured training set; the second model pre-training unit is used for training the preset training corpus model through the second pre-training set so as to obtain a second pre-trained model; the second model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the second pre-trained model through the second fine tuning training set so as to obtain a second labeling model; and the medical document structured database generation unit is used for extracting the time tag of the attribute data in the medical document structured training set, and processing the medical document data in the medical text database based on the time tag and the second annotation model to obtain the medical document structured database.
After the structured training set is obtained, the structured training set needs to be split into the pre-training set D1 and the fine-tuning training set D2, as shown in fig. 3 and 4, the data in the fine-tuning training set D2 needs to be classified based on the designated clinical feature attribute, and the pre-training set D1 is not processed. After the pre-training set D1 and the fine tuning training set D2 are determined, the pre-training set D1 is required to be used as training data to train the pre-training corpus model through a pre-training method to obtain a pre-training model M1, the pre-training method includes, but is not limited to, BERT (Bidirectional Encoder Representation from Transformer) training, and the classification attribute of the fine tuning training set D2 includes, but is not limited to, a disease type, a treatment plan type, and a combination of the disease type and the treatment plan type. After the pre-training model M1 is obtained, the pre-training model M1 needs to be subjected to fine-tuning training, the pre-training model M1 can be subjected to fine-tuning training for one time by data corresponding to each disease feature type in the fine-tuning training set D2, in order to save hardware resources during later deployment in each fine-tuning training process, in this embodiment, a whole set of models cannot be stored for each disease feature independently, but the structure and parameters before the full-connection layer of the pre-training model M1 are kept unchanged, the full-connection layer is subjected to parameter tuning by using the fine-tuning training set D2, and the full-connection layer parameters corresponding to the current disease feature are stored after fine-tuning of each disease feature data is completed. After the pretraining model M1 is successfully fine-tuned, a labeling model M2 may be generated according to a full connection layer corresponding to the current disease feature in a part before the full connection layer of the pretraining model M1, and then, labeling processing is performed on the medical paper data and the medical document data in the medical text database through the labeling model M2, so as to create a medical paper structured database and a medical document structured database respectively.
It should be further noted that, when the labeling model M2 is used for labeling the medical paper data in the medical text database, the labeling model M2 is used for automatically labeling the entity type and entity relationship of the medical paper data, so as to create a medical paper structured database through the labeled medical paper data; when the medical document data in the medical text database is marked by the marking model M2, the entity type and entity relation of the medical document data are automatically marked by the marking model M2, after the automatic marking entity and relation are obtained, the time label corresponding to the medical document data is extracted according to the preset purposes of clinical data mining and clinical outcome prediction, wherein the time label comprises, but is not limited to, the first time of patient admission, the time of patient physiological characteristic evaluation, the time of patient disease diagnosis, the time of treatment scheme implementation, the occurrence time of adverse reaction, the follow-up record time and the treatment effect evaluation time, the time label is used as a double label of the medical document data, so that the final medical document data after marking is determined, and the medical document structured database is created by the medical document data after marking.
Still another embodiment of the present invention is directed to the medical clinical outcome prediction system generating apparatus, further comprising: the model training module is used for training a preset model to be trained based on the historical training set so as to obtain a preset training corpus model. The preset model to be trained, namely the corpus model, can be trained through a historical training set generated by the historical medical corpus data, and the preset training corpus model is obtained.
In this way, in the process of constructing the structured database of the medical paper and the structured database of the medical document, in order to improve the labeling performance of the model, the training set can be divided into a pre-training set and a fine-tuning training set according to the appointed clinical characteristic attribute, and the model can be enabled to be called to analyze the new text method by the corresponding model based on the appointed clinical characteristic attribute of the text to be labeled in the application process through the adjustment of the fine-tuning training set, so that the deployment resource is optimized, and the model efficiency is improved.
The system generating module 14 is configured to perform database integration on the medical paper structured database and the medical document structured database, and generate a clinical result prediction system through the integrated fusion structured database, so as to predict a clinical result through the clinical result prediction system.
In this embodiment, the system generating module is configured to perform database integration on the medical paper structured database and the medical document structured database, and generate a clinical result prediction system through the integrated fusion structured database, so as to predict a clinical result through the clinical result prediction system. That is, the entity relationship and entity type in the abstract paper data may be extracted from the obtained medical paper structured database, a first set of entity relationships corresponding to the medical paper structured database may be generated, and the entity relationship and entity type in the abstract paper data may be extracted from the obtained medical document structured database, and a second set of entity relationships corresponding to the medical document structured database may be generated. And then carrying out relationship correction on the second entity relationship set through the first entity relationship set to obtain a target entity relationship set for creating the clinical result prediction system, and finally generating the clinical result prediction system based on the target entity relationship set and a preset retrieval tool.
It can be seen that, the medical text database creation module in this embodiment is configured to perform data extraction on acquired medical data, perform preprocessing on the acquired medical paper data and medical document data, and create a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data; the training set generation module is used for generating a medical paper structured training set based on the preprocessed medical paper data and generating a medical document structured training set based on the preprocessed medical document data; the structured database creation module is used for training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively so as to generate a corresponding medical paper structured database and a corresponding medical document structured database; and the system generation module is used for integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system. Therefore, the method and the system respectively preprocess the acquired medical paper data and the medical document data, generate corresponding structured training sets by utilizing the preprocessed data, train the model by utilizing the structured training sets, generate a medical paper structured database and a medical document structured database, and finally integrate the medical paper structured database and the medical document structured database into a database and construct a clinical result prediction system. Thus, after the literature data in the medical paper data are extracted, the clinical result can be predicted by combining the authoritative literature data and the actual clinical data, and the accuracy of the clinical result prediction can be effectively improved. Moreover, since a large amount of medical paper data and medical staff daily medical document data are utilized in the process of creating the clinical result prediction system, on one hand, diagnosis suggestions and treatment plan suggestions can be provided for patients through the created clinical result prediction system, and include, but are not limited to, diagnosis and treatment suggestions in a clinical decision support system and decision support suggestions for patient hospital management; on the other hand, clinical research direction advice may be provided to the patient, including, but not limited to, research direction advice that recommends potentially better regimen studies for the patient type for which treatment regimens have been determined, clinical outcome studies such as treatment efficacy and adverse reaction assessment points for existing and potentially better regimens, and research direction advice that the relevant regimen is best suited for patient subgroup feature screening; in yet another aspect, relevant data operations and applications may be performed based on data models collected and constructed by the system, including but not limited to insurance model recommendations for patients, health management model recommendations, statistical models, and statistical result presentation model recommendations applicable for different analytical purposes.
Referring to fig. 5, an embodiment of the present invention discloses a medical clinical outcome prediction system generating device, including:
a first entity relationship set generating unit 21, configured to match preset medical document data based on the medical paper structured database, and generate a first entity relationship set according to the matched entity type and entity relationship;
a second entity relationship set generating unit 22, configured to match preset medical document data based on the medical document structured database, and generate a second entity relationship set according to the matched entity type and entity relationship;
a third entity relationship set generating unit 23, configured to perform entity type filtering and entity relationship correction on the second entity relationship set based on the first entity relationship set, so as to obtain a target entity relationship set;
the system generating unit 24 is configured to establish a fusion structured database based on the target entity relationship set, and generate a clinical result prediction system based on the fusion structured database and a preset search tool, so as to predict a clinical result by the clinical result prediction system.
In this embodiment, after the medical paper structured database and the medical document structured database are created, the medical paper structured database may be used as a first local knowledge base, and then entity type extraction and entity relationship extraction based on the first local knowledge base are performed on preset medical document data through the first local knowledge base, so as to obtain an entity relationship set ER1 corresponding to the preset medical document data; after determining the entity relation set ER1, the medical document structured database can be used as a second local knowledge base, then entity type extraction and entity relation extraction based on the second local knowledge base are carried out on preset medical document data through the second local knowledge base to obtain an entity relation set ER2 corresponding to the preset medical document data, and then the entity type and the entity relation in ER2 are filtered by referring to the entity type in ER1 to obtain a filtered entity set E; and finally, correcting the entity relationship in ER2 based on the entity relationship in the first local knowledge base, combining the corrected entity relationship set ER2, establishing a corresponding entity relationship for the entity type in E by using a mapping mode to obtain an entity relationship set ER3, and establishing a fusion structured database based on the entity relationship set ER 3. After the fusion structured database is obtained, a clinical result prediction system can be created based on the fusion structured database and a preset search tool, so as to realize data multidimensional analysis in clinical research and application and provide a clinical result prediction function.
For example, based on the medical paper data "patient inclusion exclusion criteria" and the medical document data "patient basic information, complaints, current medical history, past history, auxiliary examination, preliminary diagnosis, admission diagnosis, specialty, diagnosis pass" and the like, after integrating the relevant data in the structured database, the physiological characteristics of the patient and the disease differentiation diagnosis recommendation can be analyzed and studied; for another example, based on the medical paper data "treatment plan name-period and frequency of treatment plan-treatment effect evaluation end point type-end point result data" and the medical document data "treatment plan name-number of times treatment plan is performed and specific time-treatment end point type-treatment end point time is reached", and the like, after the correlation data in the structured database is integrated, the relationship between the adjustment of treatment plan and treatment effect can be analyzed and studied.
In this way, prediction of subdivided disease diagnosis types based on physiological characteristics of patients in the same department can be achieved through the created clinical outcome prediction system; under the condition of the same disease characteristics, the treatment outcome difference prediction caused by the treatment scheme difference is realized; achieving a prediction of treatment outcome differences for the same treatment regimen due to adjustments in the real world under the same disease signature conditions; under the condition of realizing the same treatment scheme, the treatment outcome difference of patients with different physiological characteristics is predicted; prediction of treatment outcome impact of different disease treatment histories on patients with the same disease characteristics and treatment regimen is achieved.
Referring to fig. 6, the embodiment of the invention discloses a medical clinical result prediction system generation method, which comprises the following steps:
and S11, carrying out data extraction on the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data.
And step S12, a medical paper structured training set is generated based on the preprocessed medical paper data, and a medical document structured training set is generated based on the preprocessed medical document data.
And step S13, training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively to generate a corresponding medical paper structured database and a corresponding medical document structured database.
And step S14, integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system.
Therefore, the method comprises the steps of firstly, extracting data of acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data; then, a medical paper structured training set is generated based on the preprocessed medical paper data, and a medical document structured training set is generated based on the preprocessed medical document data; training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively to generate a corresponding medical paper structured database and a corresponding medical document structured database; and finally, integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system. Therefore, the method and the system respectively preprocess the acquired medical paper data and the medical document data, generate corresponding structured training sets by utilizing the preprocessed data, train the model by utilizing the structured training sets, generate a medical paper structured database and a medical document structured database, and finally integrate the medical paper structured database and the medical document structured database into a database and construct a clinical result prediction system. Thus, after the literature data in the medical paper data are extracted, the clinical result can be predicted by combining the authoritative literature data and the actual clinical data, and the accuracy of the clinical result prediction can be effectively improved. Moreover, since a large amount of medical paper data and medical staff daily medical document data are utilized in the process of creating the clinical result prediction system, on one hand, diagnosis suggestions and treatment plan suggestions can be provided for patients through the created clinical result prediction system, and include, but are not limited to, diagnosis and treatment suggestions in a clinical decision support system and decision support suggestions for patient hospital management; on the other hand, clinical research direction advice may be provided to the patient, including, but not limited to, research direction advice that recommends potentially better regimen studies for the patient type for which treatment regimens have been determined, clinical outcome studies such as treatment efficacy and adverse reaction assessment points for existing and potentially better regimens, and research direction advice that the relevant regimen is best suited for patient subgroup feature screening; in yet another aspect, relevant data operations and applications may be performed based on data models collected and constructed by the system, including but not limited to insurance model recommendations for patients, health management model recommendations, statistical models, and statistical result presentation model recommendations applicable for different analytical purposes.
In some embodiments, the data extracting the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data includes:
acquiring medical data, extracting medical paper data and medical document data in the medical data, and respectively marking attributes of the medical paper data and the medical document data to obtain preprocessed medical paper data and preprocessed medical document data;
a medical text database is created based on the preprocessed medical paper data and the preprocessed medical document data.
In some embodiments, the generating a medical paper structured training set based on the preprocessed medical paper data and generating a medical document structured training set based on the preprocessed medical document data comprises:
acquiring the preprocessed medical paper data and the preprocessed medical document data stored in the medical text database, and respectively determining a relation between a first entity type and a first entity in the preprocessed medical paper data and a relation between a second entity type and a second entity in the medical document data;
Creating a first data tag based on the relation between the first entity type and the first entity, marking the preprocessed medical paper data through the first data tag, and taking the marked medical paper data as first training data;
creating a second data tag based on the relation between the second entity type and the second entity, marking the preprocessed medical document data through the second data tag, and taking the marked medical document data as second training data;
and respectively carrying out standardization operation and normalization operation on the first training data and the second training data to respectively generate a medical paper structured training set and a medical document structured training set.
In some embodiments, the medical clinical outcome prediction system generation method further comprises:
training the preset model to be trained based on the historical training set to obtain a preset training corpus model.
In some embodiments, the training the preset training corpus model through the medical paper structured training set and the medical document structured training set to generate a corresponding medical paper structured database and a medical document structured database, respectively, includes:
Generating a first pre-training set and a first fine tuning training set based on the medical paper structured training set;
training the preset training corpus model through the first pre-training set to obtain a first pre-trained model;
performing fine tuning training on all the connection layer parameters of the first pre-trained model through the first fine tuning training set to obtain a first labeling model;
and processing the medical paper data in the medical text database through the first labeling model to obtain a medical paper structured database.
In some embodiments, the training the preset training corpus model through the medical paper structured training set and the medical document structured training set to generate a corresponding medical paper structured database and a medical document structured database, respectively, includes:
generating a second pre-training set and a second fine tuning training set based on the medical document structured training set;
training the preset training corpus model through the second pre-training set to obtain a second pre-trained model;
performing fine tuning training on all the connection layer parameters of the second pre-trained model through the second fine tuning training set to obtain a second labeling model;
And extracting a time tag of attribute data in the medical document structured training set, and processing medical document data in the medical text database based on the time tag and the second annotation model to obtain the medical document structured database.
In some embodiments, the database integration of the medical paper structured database and the medical document structured database, and the generation of the clinical outcome prediction system by the integrated fusion structured database, to predict the clinical outcome by the clinical outcome prediction system, comprises:
matching preset medical document data based on the medical paper structured database, and generating a first entity relationship set through the matched entity type and entity relationship;
matching preset medical document data based on the medical document structured database, and generating a second entity relationship set through the matched entity type and entity relationship;
performing entity type filtering and entity relationship correction on the second entity relationship set based on the first entity relationship set to obtain a target entity relationship set;
and establishing a fusion structured database based on the target entity relation set, and generating a clinical result prediction system based on the fusion structured database and a preset retrieval tool so as to predict a clinical result through the clinical result prediction system.
Further, the embodiment of the present application further discloses an electronic device, and fig. 7 is a block diagram of an electronic device 200 according to an exemplary embodiment, where the content of the figure is not to be considered as any limitation on the scope of use of the present application.
Fig. 7 is a schematic structural diagram of an electronic device 200 according to an embodiment of the present application. The electronic device 200 may specifically include: at least one processor 201, at least one memory 202, a power supply 203, a communication interface 204, an input output interface 205, and a communication bus 206. Wherein the memory 202 is used for storing a computer program, which is loaded and executed by the processor 201 to implement the relevant steps in the medical clinical outcome prediction system generation method disclosed in any of the foregoing embodiments. In addition, the electronic apparatus 200 in the present embodiment may be specifically an electronic computer.
In this embodiment, the power supply 203 is configured to provide an operating voltage for each hardware device on the electronic device 200; the communication interface 204 can create a data transmission channel between the electronic device 200 and an external device, and the communication protocol to be followed by the communication interface is any communication protocol applicable to the technical solution of the present application, which is not specifically limited herein; the input/output interface 205 is used for obtaining external input data or outputting external output data, and the specific interface type thereof may be selected according to the specific application requirement, which is not limited herein.
The memory 202 may be a carrier for storing resources, such as a read-only memory, a random access memory, a magnetic disk, or an optical disk, and the resources stored thereon may include an operating system 221, a computer program 222, and the like, and the storage may be temporary storage or permanent storage.
The operating system 221 is used for managing and controlling various hardware devices on the electronic device 200 and the computer program 222, which may be Windows Server, netware, unix, linux, etc. The computer program 222 may further comprise a computer program capable of being used to perform other specific tasks in addition to the computer program capable of being used to perform the medical clinical outcome prediction system generation method performed by the electronic device 200 as disclosed in any of the previous embodiments.
Further, the application also discloses a computer readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the medical clinical outcome prediction system generation method of the foregoing disclosure. For specific steps of the method, reference may be made to the corresponding contents disclosed in the foregoing embodiments, and no further description is given here.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, so that the same or similar parts between the embodiments are referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. 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.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing has outlined rather broadly the more detailed description of the application in order that the detailed description of the application that follows may be better understood, and in order that the present principles and embodiments may be better understood; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.
Claims (10)
1. A medical clinical outcome prediction system generation device, comprising:
the medical text database creation module is used for carrying out data extraction on the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data;
the training set generation module is used for generating a medical paper structured training set based on the preprocessed medical paper data and generating a medical document structured training set based on the preprocessed medical document data;
the structured database creation module is used for training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively so as to generate a corresponding medical paper structured database and a corresponding medical document structured database;
And the system generation module is used for integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system.
2. The medical clinical outcome prediction system generation apparatus according to claim 1, wherein the medical text database creation module comprises:
the text data preprocessing unit is used for acquiring medical data, extracting medical paper data and medical document data in the medical data, and respectively marking attributes of the medical paper data and the medical document data to obtain preprocessed medical paper data and preprocessed medical document data;
and the database creation unit is used for creating a medical text database based on the preprocessed medical paper data and the preprocessed medical document data.
3. The medical clinical outcome prediction system generation apparatus according to any one of claims 1 or 2, wherein the training set generation module includes:
The data attribute determining unit is used for acquiring the preprocessed medical paper data and the preprocessed medical document data stored in the medical text database and respectively determining a relation between a first entity type and a first entity in the preprocessed medical paper data and a relation between a second entity type and a second entity in the medical document data;
the first training data determining unit is used for creating a first data tag based on the relation between the first entity type and the first entity, marking the preprocessed medical paper data through the first data tag, and taking the marked medical paper data as first training data;
the second training data determining unit is used for creating a second data tag based on the relation between the second entity type and the second entity, marking the preprocessed medical document data through the second data tag, and taking the marked medical document data as second training data;
and the training set generating unit is used for respectively carrying out standardization operation and normalization operation on the first training data and the second training data so as to respectively generate a medical paper structured training set and a medical document structured training set.
4. The medical clinical outcome prediction system generating device according to claim 1, further comprising:
the model training module is used for training a preset model to be trained based on the historical training set so as to obtain a preset training corpus model.
5. The medical clinical outcome prediction system generation apparatus according to claim 2, wherein the structured database creation module comprises:
a first model training set generating unit, configured to generate a first pre-training set and a first fine tuning training set based on the medical paper structured training set;
the first model pre-training unit is used for training the preset training corpus model through the first pre-training set so as to obtain a first pre-trained model;
the first model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the first pre-trained model through the first fine tuning training set so as to obtain a first labeling model;
and the medical paper structured database generation unit is used for processing the medical paper data in the medical text database through the first labeling model so as to obtain the medical paper structured database.
6. The medical clinical outcome prediction system generation apparatus according to claim 2, wherein the structured database creation module comprises:
a second model training set generating unit, configured to generate a second pre-training set and a second fine tuning training set based on the medical document structured training set;
the second model pre-training unit is used for training the preset training corpus model through the second pre-training set so as to obtain a second pre-trained model;
the second model fine tuning unit is used for carrying out fine tuning training on the full-connection layer parameters of the second pre-trained model through the second fine tuning training set so as to obtain a second labeling model;
and the medical document structured database generation unit is used for extracting the time tag of the attribute data in the medical document structured training set, and processing the medical document data in the medical text database based on the time tag and the second annotation model to obtain the medical document structured database.
7. The medical clinical outcome prediction system generation apparatus of claim 6, wherein the system generation module comprises:
The first entity relation set generation unit is used for matching preset medical document data based on the medical paper structured database and generating a first entity relation set through the matched entity type and entity relation;
the second entity relationship set generating unit is used for matching preset medical document data based on the medical document structured database and generating a second entity relationship set through the matched entity type and entity relationship;
a third entity relationship set generating unit, configured to perform entity type filtering and entity relationship correction on the second entity relationship set based on the first entity relationship set, so as to obtain a target entity relationship set;
and the system generation unit is used for establishing a fusion structured database based on the target entity relation set, and generating a clinical result prediction system based on the fusion structured database and a preset retrieval tool so as to predict clinical results through the clinical result prediction system.
8. A method for generating a medical clinical outcome prediction system, comprising:
extracting data from the acquired medical data, preprocessing the acquired medical paper data and medical document data, and creating a medical text database based on the acquired preprocessed medical paper data and preprocessed medical document data;
Generating a medical paper structured training set based on the preprocessed medical paper data, and generating a medical document structured training set based on the preprocessed medical document data;
training a preset training corpus model through the medical paper structured training set and the medical document structured training set respectively to generate a corresponding medical paper structured database and a corresponding medical document structured database;
and integrating the medical paper structured database and the medical document structured database, and generating a clinical result prediction system through the integrated fusion structured database so as to predict clinical results through the clinical result prediction system.
9. An electronic device, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the medical clinical outcome prediction system generation method of claim 8.
10. A computer readable storage medium storing a computer program which when executed by a processor implements the medical clinical outcome prediction system generation method of claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311045311.5A CN117059216A (en) | 2023-08-18 | 2023-08-18 | Medical clinical result prediction system generation device, method, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311045311.5A CN117059216A (en) | 2023-08-18 | 2023-08-18 | Medical clinical result prediction system generation device, method, equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117059216A true CN117059216A (en) | 2023-11-14 |
Family
ID=88668883
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311045311.5A Pending CN117059216A (en) | 2023-08-18 | 2023-08-18 | Medical clinical result prediction system generation device, method, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117059216A (en) |
-
2023
- 2023-08-18 CN CN202311045311.5A patent/CN117059216A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20130311201A1 (en) | Medical record generation and processing | |
| US11406265B2 (en) | Method for automating collection, association, and coordination of multiple medical data sources | |
| CN109584975B (en) | Medical data standardization processing method and device | |
| US8612261B1 (en) | Automated learning for medical data processing system | |
| US20160210426A1 (en) | Method of classifying medical documents | |
| KR100538577B1 (en) | Method For Standardization Of Computerization Of Medical Information | |
| US10679738B2 (en) | Identification of codable sections in medical documents | |
| JP5952835B2 (en) | Imaging protocol updates and / or recommenders | |
| US8312057B2 (en) | Methods and system to generate data associated with a medical report using voice inputs | |
| CN108899064A (en) | Electronic health record generation method, device, computer equipment and storage medium | |
| US20060136259A1 (en) | Multi-dimensional analysis of medical data | |
| US20050055246A1 (en) | Patient workflow process | |
| US20070237378A1 (en) | Multi-input reporting and editing tool | |
| EP1573624A4 (en) | Systems and methods for automated extraction and processing of billing information in patient records | |
| WO2019070200A1 (en) | A healthcare information management system | |
| CN111081329A (en) | Automatic clinical data entry method and device, electronic equipment and storage medium | |
| CN111723277A (en) | Image structured report cloud requirement classification system and method thereof | |
| US11200967B1 (en) | Medical patient synergistic treatment application | |
| Varney et al. | The potential for using ChatGPT to improve imaging appropriateness | |
| US20200381108A1 (en) | System and Method for Managing Diagnostic Imaging Costs | |
| US20010032102A1 (en) | Psychiatric information systems, methods and computer program products that capture psychiatric information as discrete data elements | |
| JP7278256B2 (en) | Devices, systems and methods for optimizing image acquisition workflows | |
| US20230317278A1 (en) | System for Aggregating, Analyzing, and Reporting Medical Information | |
| CN117059216A (en) | Medical clinical result prediction system generation device, method, equipment and medium | |
| WO2021002847A1 (en) | Method for automating collection, association, and coordination of multiple medical data sources |
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 |