CN112466462B - EMR information association and evolution method based on deep learning of image - Google Patents
EMR information association and evolution method based on deep learning of image Download PDFInfo
- Publication number
- CN112466462B CN112466462B CN202011353177.1A CN202011353177A CN112466462B CN 112466462 B CN112466462 B CN 112466462B CN 202011353177 A CN202011353177 A CN 202011353177A CN 112466462 B CN112466462 B CN 112466462B
- Authority
- CN
- China
- Prior art keywords
- emr
- deep learning
- map
- data
- image
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- 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
-
- 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
- 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
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- 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 invention discloses an EMR information association and evolution method based on deep learning of an image, which comprises the following steps: EMR data preprocessing: acquiring an EMR data set to obtain an entity dictionary of EMR; EMR map construction: converting words in the entity dictionary into vector representation, obtaining a vector matrix of the EMR diagram and an adjacent matrix of the EMR diagram, and combining to form the EMR diagram; EMR image deep learning model construction: constructing an input image data set of the EMR image deep learning model according to all EMR images corresponding to the obtained EMR data set, thereby further obtaining an image deep learning model F; EMR information association and evolution: and feeding any piece of EMR data into the image deep learning model F to construct an evolution sequence of the EMR image. The invention provides an EMR information association and evolution method based on graph deep learning.
Description
Technical Field
The invention relates to the field of natural language processing and deep learning, in particular to an EMR information association and evolution method based on image deep learning.
Background
Clinical texts contain abundant health and medical information, and are an important information resource generated during medical activities, wherein the clinical texts are represented by Electronic Medical Records (EMRs). The electronic medical records are divided into outpatient electronic medical records and inpatient electronic medical records and comprise social demographic information, chief complaints, current medical history, examination records, disease diagnosis and the like. As can be seen, the electronic medical record is a multi-source heterogeneous data set with high medical knowledge density, and contains abundant entities, such as: symptoms, diseases, examinations, etc., there is often some kind of medical relationship hidden between these entities. At present, effective modeling of electronic medical records has become an important issue in academia and industry. Research has shown that machine learning using electronic medical record data can realize intelligent clinical applications such as disease diagnosis, drug recommendation, treatment plan recommendation, risk prediction, etc. However, most electronic medical records are stored in an unstructured text form, which results in low application efficiency of the medical records, and hinders the degree of medical informatization, and clinical workers cannot clearly acquire the structured associated information and medical knowledge of the patient's condition. How to find clinical knowledge from massive electronic medical record data is a challenge in the health and medical field, and is also an important way to improve the efficiency of medical scientific research and to seek reliable evidence of clinical diagnosis.
Disclosure of Invention
The invention mainly aims to overcome the defects in the prior art and provides an EMR information association and evolution method based on graph deep learning.
The invention adopts the following technical scheme:
an EMR information association and evolution method based on image deep learning is characterized by comprising the following steps:
EMR data preprocessing: acquiring an EMR data set, and preprocessing the acquired EMR data to obtain an entity dictionary of EMR;
EMR map construction: converting words in the entity dictionary into vector representation by using a word2vec method to obtain a vector matrix of an EMR map corresponding to the EMR data; sorting machineTaking entity words of the entity dictionary as nodes of the EMR map, wherein the number of the entity words in the entity dictionary is the maximum value V _ num of the number of the nodes of the EMR map, and obtaining an adjacency matrix of the EMR map by calculating the conditional probability of any two map nodes and normalizing the adjacency matrix to obtain the adjacency matrix of the EMR map; the vector matrix of the EMR map and the adjacent matrix of the EMR map form the EMR map; wherein, aiming at any two nodes v in the EMR diagram i And v j (i, j =1,2, ·, V _ num, and i ≠ j), V j To v i The edge weight of is P (v) i |v j ) I.e. v i At v j The probability of occurrence under the occurrence condition is calculated by the formula:
EMR image deep learning model construction: constructing an input graph data set T = { G ] of an EMR graph deep learning model according to all EMR graphs corresponding to the obtained EMR data set i I =1,2, …, n }, where G i =(Q i ,A i ) Is the ith electronic medical record data, n is the number of the electronic medical record data, Q i ={q_vecoter i I =1,2, …, V _ num } is EMR map G i V _ num is the maximum value of the node number of the EMR graph, A i I =1,2, …, n is EMR map G i The adjacency matrix of (a); performing EMR image deep learning by using graph neural network transducer, and enabling adjacent matrix A i First self-attention Module Block as a neural network transducer 1 Of the initialization matrix M 1 I.e. A i =M 1 (ii) a Adopting the graph data set T as input data of a graph neural network transform model, taking the primary diagnosis of EMR as output data of the graph neural network transform model, and training the transform model to obtain a graph deep learning model F;
EMR information association and evolution: feeding any piece of EMR image data into an image deep learning model F, and passing through a second and more self-attention modules block of the model F i (i =2, …, M) to construct a series of contiguous matrices M of EMR maps G 2 ,…,M m Wherein M is 2 ,…,M m Is formed by a adjacency matrix M based on conditional probability 1 Evolution of adjacency matrixes obtained through deep learning of the graph; collection M 1 ,M 2 ,…,M m Evolution sequence M = { M) for constructing EMR map i ,i=1,2,…,m}。
Specifically, the EMR data sets are acquired, including medical texts such as social demographic information, chief complaints, physical examinations, laboratory examinations, and results and disease diagnoses thereof.
Specifically, the pre-processing of the acquired EMR data includes word segmentation and entity extraction.
As can be seen from the above description of the present invention, compared with the prior art, the present invention has the following advantages:
the graph structure information of the EMR is used for representation learning, and a large amount of associated information of medical knowledge contained in the electronic medical record data and an evolution rule thereof can be mined. In one aspect, the graph-structured electronic medical record data can yield valuable information and knowledge to provide clinical decision support for physicians. On the other hand, the evolution visualization technology of the graph data can be used for displaying the association change process of the electronic medical record information, so that deep learning has interpretability, and the practical application of medical artificial intelligence is better served.
Drawings
FIG. 1 is a flow chart of EMR information association and evolution method based on graph deep learning according to the invention;
FIG. 2 is an exemplary diagram of Chinese electronic medical record data for pediatric clinics;
FIG. 3 is a diagram of an EMR map deep learning model;
FIG. 4 is a schematic diagram of EMR chart structure information input;
FIG. 5 is a schematic diagram of the evolution of EMR map structural information;
fig. 6 is a schematic diagram of structural information output of an EMR map.
The invention is described in further detail below with reference to the figures and specific examples.
Detailed Description
The invention will be further illustrated with reference to the following specific examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. Further, it should be understood that various changes or modifications of the present invention may be made by those skilled in the art after reading the teaching of the present invention, and such equivalents may fall within the scope of the present invention as defined in the appended claims.
The figure is a representation method for naturally and visually displaying object association, and is the most common information carrier and form in our production and life. Graph-based representation learning research aims at better analyzing the connections between nodes in a complex information network and their evolution process. The EMR map can be constructed based on the association information of the objective data storage of the electronic medical record or based on the statistical information of multiple layers and angles. The graph representation-based method can effectively learn the potential structural information of the text of the electronic medical record and the semantic relation rich in data. Researches show that high-quality vector representation can be obtained by using graph structure information of electronic medical record text data, and the performance of downstream tasks is remarkably improved. Based on the EMR information association and evolution method, the EMR information association and evolution method based on the deep learning of the image is provided.
Referring to fig. 1, the EMR information association and evolution method based on graph deep learning of the present invention includes the following steps: (1) EMR data preprocessing; (2) EMR map construction; (3) EMR image deep learning model construction; and (4) EMR information association and evolution.
Specifically, EMR data preprocessing: acquiring an EMR data set which comprises medical texts such as social and demographic information and chief complaints, physical examination, laboratory examination and results thereof and disease diagnosis; preprocessing the acquired EMR data, including word segmentation and entity extraction to obtain an entity dictionary of the EMR;
EMR map construction: converting words in the entity dictionary into vector representation by using a word2vec method to obtain a vector matrix of an EMR map corresponding to the EMR data; selecting entity words of an entity dictionary as nodes of the EMR graph, wherein the number of the entity words in the entity dictionary is the maximum value of the number of the nodes of the EMR graph, and obtaining an adjacent matrix of the EMR graph and normalizing the adjacent matrix by calculating the conditional probability of any two graph nodes to obtain the adjacent matrix of the EMR graph(ii) a The vector matrix of the EMR map and the adjacent matrix of the EMR map form the EMR map; wherein, aiming at any two nodes v in the EMR diagram i And v j (i, j =1,2, ·, V _ num, and i ≠ j), V j To v i The edge weight of is P (v) i |v j ) I.e. v i At v j The probability of occurrence under the occurrence condition is calculated by the formula:
EMR image deep learning model construction: constructing an input graph data set T = { G ] of an EMR graph deep learning model according to all EMR graphs corresponding to the obtained EMR data set i I =1,2, …, n }, where G i =(Q i ,A i ) Is the ith electronic medical record data, n is the number of the electronic medical record data, Q i ={q_vecoter i I =1,2, …, V _ num } is EMR map G i V _ num is the maximum value A of the node number of the EMR graph i I =1,2, …, n is EMR map G i The adjacency matrix of (a); performing EMR image deep learning by using graph neural network transducer, and enabling adjacent matrix A i First self-attention Module Block as a neural network transducer 1 Of the initialization matrix M 1 I.e. A i =M 1 (ii) a Adopting the graph data set T as input data of a graph neural network transform model, taking the primary diagnosis of EMR as output data of the graph neural network transform model, and training the transform model to obtain a graph deep learning model F;
EMR information association and evolution: feeding any piece of EMR image data into an image deep learning model F, and passing through a second and more self-attention modules block of the model F i (i =2, …, M) to construct a contiguous matrix M of a series of EMR maps G 2 ,…,M m Wherein M is 2 ,…,M m Is formed by an adjacency matrix M based on conditional probability 1 Evolution of the adjacency matrix obtained through deep learning of the graph; collection M 1 ,M 2 ,…,M m Evolving sequence M for constructing EMR map={M i ,i=1,2,…,m}。
Taking an electronic medical record from a certain hospital, fig. 2 is an example of a real chinese electronic medical record data for pediatric outpatient service, which includes basic information of patients, chief complaints, current medical history, past medical history, family medical history, physical examination, auxiliary examination results, and preliminary diagnosis, and in which there are a large number of medical professional terms.
The embodiment of the invention comprises the following specific steps:
the method comprises the following steps: EMR data preprocessing.
Firstly, the electronic medical record data needs to be accurately segmented, and the quality of the segmentation affects the text mining effect. In the word segmentation stage, the invention combines a self-defined medical dictionary and adopts a word segmentation tool to segment the electronic medical record text. In the entity extraction stage, meaningful entities are further extracted from the segmented results, and finally, each piece of unstructured electronic medical record data is converted into a structured entity word list. Through the above operations, the dictionary Dict of all electronic medical record data can be obtained, and the scale is 12310. And further generating a 128-dimensional word vector expression q _ vector for each entity word q e Dict by using a word2vec method.
Step two: and (4) EMR map construction.
And selecting the entity word q belonging to Dict of each piece of electronic medical record data as a node V of the EMR graph G, wherein the maximum value V _ num of the number of the nodes of the electronic medical record is 150. Calculating any two graph nodes v i And v j (i, j =1,2., 150, and i ≠ j), acquiring an adjacency matrix of the EMR map G and normalizing the adjacency matrix to obtain A 150*150 . Finally, construction G = (Q, a) of the EMR graph is completed, where Q is a vector representation of the nodes and a is an adjacency matrix of the nodes.
Step three: and (5) constructing an EMR image deep learning model.
An EMR map deep learning model map is shown in fig. 3. Firstly, graph data of the electronic medical record, namely a node vector Q and an adjacency matrix A are used as input data of a graph neural network transform model, and the adjacency matrix A is used as a graph neural network transform model self-attention module block 1 Is initialized to matrix M 1 . Will be provided withThe preliminary diagnosis of the electronic medical record is used as output data of the graph neural network transform model. The example uses 144,170 true and valid electronic medical record image data sets T to train the transformer model to obtain an image deep learning model F.
Step four: EMR information association and evolution.
Feeding any piece of electronic medical record data into a trained EMR image deep learning model F, and extracting a self-attention module block of the second and more than one model F i (i =2, …, M) to construct a series of contiguous matrices M of EMR maps G 2 ,…,M m Wherein M is 2 ,…,M m Is formed by a adjacency matrix M based on conditional probability 1 And (4) evolution of the adjacency matrix obtained through deep learning of the graph. Collection M 1 ,M 2 ,…,M m Constructing an evolution sequence M = { M of EMR map G i I =1,2, …, m }. In this example, m =3. Fig. 4, 5, 6 show the structured information association and evolution based on the deep learning of the graph of a piece of EMR data preliminarily diagnosed as "common cold", wherein the display threshold of the edge weight of the graph is set to 0.054.
Fig. 4 shows graph data of the electronic medical record obtained based on the conditional probability, where the graph data calculated by the conditional probability is fully connected. Fig. 5 shows nodes highlighted in the data of the electronic medical record map in the deep learning process of the map, which correspond to important EMR information, including "chief complaints", "fever", "cough", "medical history", "male", "neck", "abdomen", "heart rate", "softness", and the like. Fig. 6 is a final information association and evolution result of the electronic medical record data after deep learning of the image, and shows that the key information of the electronic medical record is as follows: the sex of the patient is male, the patient has the medical history of diabetes and febrile convulsion, and the current state of the patient is 'normal' judged by 'physical examination', 'oral cavity', 'mass inclusion', 'no-rale', 'normal', 'abdomen', 'family history' and the like, but the patient has symptoms of 'no swelling-fever' and the like.
So far, EMR information association based on graph deep learning and its evolution method are all ended. As can be easily found, the EMR map data is constructed by performing natural language processing on unstructured electronic medical record data and based on objective data statistical information of the electronic medical record data. Knowledge discovery and interpretable deep learning are realized through the graph deep learning method, and clinical decision support is provided for doctors, so that the method can better serve the practical application of medical artificial intelligence.
The above description is only an embodiment of the present invention, but the design concept of the present invention is not limited thereto, and any insubstantial modifications made by using the design concept should fall within the scope of infringing the present invention.
Claims (3)
1. An EMR information association and evolution method based on deep learning of an image is characterized by comprising the following steps:
EMR data preprocessing: acquiring an EMR data set, and preprocessing the acquired EMR data to obtain an entity dictionary of the EMR;
EMR map construction: converting words in the entity dictionary into vector representation by using a word2vec method to obtain a vector matrix of an EMR map corresponding to the EMR data; selecting entity words of an entity dictionary as nodes of the EMR map, wherein the number of the entity words in the entity dictionary is the maximum value V _ num of the number of the nodes of the EMR map, and acquiring an adjacency matrix of the EMR map by calculating the conditional probability of any two map nodes and normalizing the adjacency matrix to obtain a normalized adjacency matrix of the EMR map; the vector matrix of the EMR image and the adjacent matrix of the normalized EMR image form an EMR image; wherein, aiming at any two nodes v in the EMR diagram i And v j I, j =1,2,., V _ num, and i ≠ j, V j To v i The edge weight of is P (v) i |v j ) I.e. v i At v is j The probability of occurrence under the occurrence condition is calculated by the formula:
EMR image deep learning model construction: constructing EMR image depth according to all EMR images corresponding to the obtained EMR data setInput graph data set T = { G) of degree learning model i I =1,2, …, n }, where G i =(Q i ,A i ) Is the ith electronic medical record data, n is the number of the electronic medical record data, Q i ={q_vecoter i I =1,2, …, V _ num } is EMR map G i Vector matrix of (A) i I =1,2, …, n is EMR map G i The adjacency matrix of (a); EMR image deep learning is carried out by using a graph neural network transformer model, and an adjacent matrix A is obtained i First self-attention Module Block as a neural network transducer model 1 Of the initialization matrix M 1 I.e. A i =M 1 (ii) a Adopting the graph data set T as input data of a graph neural network transform model, taking the primary diagnosis of EMR as output data of the graph neural network transform model, and training the transform model to obtain a graph deep learning model F;
EMR information association and evolution: feeding any piece of EMR image data into an image deep learning model F, and passing through a second and more self-attention modules block of the model F i To construct a series of EMR maps G i Of the adjacency matrix M 2 ,…M m Wherein M is 2 ,…M m Is formed by an adjacency matrix M based on conditional probability 1 Evolution of the adjacency matrix obtained through deep learning of the graph; collection M 1 ,M 2 ,…M m Evolution sequence M = { M) for constructing EMR map i ,i=1,2,…,m}。
2. The map deep learning-based EMR information correlation and evolution method of claim 1, wherein the EMR data sets are obtained, and include social demographic information, chief medical texts, physical examinations, laboratory examinations and their results and disease diagnoses.
3. The map deep learning-based EMR information correlation and evolution method of claim 1, wherein the pre-processing is performed on the obtained EMR data, and the pre-processing includes word segmentation and entity extraction.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011353177.1A CN112466462B (en) | 2020-11-26 | 2020-11-26 | EMR information association and evolution method based on deep learning of image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011353177.1A CN112466462B (en) | 2020-11-26 | 2020-11-26 | EMR information association and evolution method based on deep learning of image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112466462A CN112466462A (en) | 2021-03-09 |
CN112466462B true CN112466462B (en) | 2023-03-07 |
Family
ID=74808674
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011353177.1A Active CN112466462B (en) | 2020-11-26 | 2020-11-26 | EMR information association and evolution method based on deep learning of image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112466462B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113313174A (en) * | 2021-06-01 | 2021-08-27 | 北京大数医达科技有限公司 | Information display method and terminal equipment |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106484674A (en) * | 2016-09-20 | 2017-03-08 | 北京工业大学 | A kind of Chinese electronic health record concept extraction method based on deep learning |
EP3221811A1 (en) * | 2014-11-20 | 2017-09-27 | T&W Engineering A/S | Method and system for establishing network connection to a wearable eeg monitoring module |
CN109243616A (en) * | 2018-06-29 | 2019-01-18 | 东华大学 | Mammary gland electronic health record joint Relation extraction and architectural system based on deep learning |
EP3511941A1 (en) * | 2018-01-12 | 2019-07-17 | Siemens Healthcare GmbH | Method and system for evaluating medical examination results of a patient, computer program and electronically readable storage medium |
EP3537448A1 (en) * | 2018-03-06 | 2019-09-11 | Digital Surgery Ltd | Methods and systems for using multiple data structures to process surgical data |
CN111316281A (en) * | 2017-07-26 | 2020-06-19 | 舒辅医疗 | Semantic classification of numerical data in natural language context based on machine learning |
CN111737415A (en) * | 2020-06-12 | 2020-10-02 | 清华大学 | Entity relationship extraction method, and method and device for acquiring entity relationship learning model |
CN111914562A (en) * | 2020-08-21 | 2020-11-10 | 腾讯科技(深圳)有限公司 | Electronic information analysis method, device, equipment and readable storage medium |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11488714B2 (en) * | 2016-03-23 | 2022-11-01 | HealthPals, Inc. | Machine learning for collaborative medical data metrics |
CN107193919A (en) * | 2017-05-15 | 2017-09-22 | 清华大学深圳研究生院 | The search method and system of a kind of electronic health record |
CN107863147B (en) * | 2017-10-24 | 2021-03-16 | 清华大学 | Medical diagnosis method based on deep convolutional neural network |
RU2720363C2 (en) * | 2017-12-29 | 2020-04-29 | Общество С Ограниченной Ответственностью "Интеллоджик" | Method for generating mathematical models of a patient using artificial intelligence techniques |
JP2020042810A (en) * | 2018-09-06 | 2020-03-19 | キヤノンメディカルシステムズ株式会社 | Medical information processing device, medical information processing method, medical information processing program and medical information processing system |
BR112021008038A2 (en) * | 2018-11-08 | 2021-07-27 | Netflix, Inc. | identify image aesthetics using region composite graphics |
US10485489B1 (en) * | 2018-12-11 | 2019-11-26 | Fifth Eye Inc. | System and method for assessing and monitoring the hemodynamic condition of a patient |
CN110298383B (en) * | 2019-05-28 | 2021-07-13 | 中国科学院计算技术研究所 | Multi-mode deep learning-based pathology classification method and system |
CN110457682B (en) * | 2019-07-11 | 2022-08-09 | 新华三大数据技术有限公司 | Part-of-speech tagging method for electronic medical record, model training method and related device |
-
2020
- 2020-11-26 CN CN202011353177.1A patent/CN112466462B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3221811A1 (en) * | 2014-11-20 | 2017-09-27 | T&W Engineering A/S | Method and system for establishing network connection to a wearable eeg monitoring module |
CN106484674A (en) * | 2016-09-20 | 2017-03-08 | 北京工业大学 | A kind of Chinese electronic health record concept extraction method based on deep learning |
CN111316281A (en) * | 2017-07-26 | 2020-06-19 | 舒辅医疗 | Semantic classification of numerical data in natural language context based on machine learning |
EP3511941A1 (en) * | 2018-01-12 | 2019-07-17 | Siemens Healthcare GmbH | Method and system for evaluating medical examination results of a patient, computer program and electronically readable storage medium |
EP3537448A1 (en) * | 2018-03-06 | 2019-09-11 | Digital Surgery Ltd | Methods and systems for using multiple data structures to process surgical data |
CN109243616A (en) * | 2018-06-29 | 2019-01-18 | 东华大学 | Mammary gland electronic health record joint Relation extraction and architectural system based on deep learning |
CN111737415A (en) * | 2020-06-12 | 2020-10-02 | 清华大学 | Entity relationship extraction method, and method and device for acquiring entity relationship learning model |
CN111914562A (en) * | 2020-08-21 | 2020-11-10 | 腾讯科技(深圳)有限公司 | Electronic information analysis method, device, equipment and readable storage medium |
Non-Patent Citations (2)
Title |
---|
"A Novel Knowledge Base Question Answering Model Based on Knowledge Representation and Recurrent Convolutional Neural Network";Changhao Liu;《IEEE》;20200826;全文 * |
"Generation of Synthetic Electronic Medical Record Text";jia qi Guan etc;《IEEE》;20181206;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN112466462A (en) | 2021-03-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109920501B (en) | Electronic medical record classification method and system based on convolutional neural network and active learning | |
Zheng et al. | An automatic diagnosis of arrhythmias using a combination of CNN and LSTM technology | |
WO2023202508A1 (en) | Cognitive graph-based general practice patient personalized diagnosis and treatment scheme recommendation system | |
Shin et al. | Classification of radiology reports using neural attention models | |
Niu et al. | A deep-learning approach to ECG classification based on adversarial domain adaptation | |
JP7464800B2 (en) | METHOD AND SYSTEM FOR RECOGNITION OF MEDICAL EVENTS UNDER SMALL SAMPLE WEAKLY LABELING CONDITIONS - Patent application | |
CN116682553A (en) | Diagnosis recommendation system integrating knowledge and patient representation | |
Hammad et al. | Automated detection of myocardial infarction and heart conduction disorders based on feature selection and a deep learning model | |
Mohanty et al. | Using deep learning architectures for detection and classification of diabetic retinopathy | |
Lei et al. | An improved UNet++ model for congestive heart failure diagnosis using short-term RR intervals | |
Papandrianos et al. | An explainable classification method of SPECT myocardial perfusion images in nuclear cardiology using deep learning and grad-CAM | |
Pendyala et al. | Automated medical diagnosis from clinical data | |
CN112562808A (en) | Patient portrait generation method and device, electronic equipment and storage medium | |
Li et al. | Classification of body constitution based on TCM philosophy and deep learning | |
Alalayah et al. | Automatic and early detection of Parkinson’s disease by analyzing acoustic signals using classification algorithms based on recursive feature elimination method | |
WO2019132686A1 (en) | Method for generating mathematical models of a patient using artificial intelligence technologies | |
CN112466462B (en) | EMR information association and evolution method based on deep learning of image | |
Shastry et al. | Deep learning techniques for the effective prediction of Alzheimer’s disease: a comprehensive review | |
Leng et al. | Bi-level artificial intelligence model for risk classification of acute respiratory diseases based on Chinese clinical data | |
Rajesh et al. | Machine learning for online automatic prediction of common disease attributes using never-ending image learner | |
Ju et al. | 3D-CNN-SPP: A patient risk prediction system from electronic health records via 3D CNN and spatial pyramid pooling | |
Huang et al. | DSCNN-LSTMs: A lightweight and efficient model for epilepsy recognition | |
CN117316466A (en) | Clinical decision method, system and equipment based on knowledge graph and natural language processing technology | |
Aguiar-Salazar et al. | Rapid detection of cardiac pathologies by neural networks using ECG signals (1D) and sECG images (3D) | |
Hu et al. | Psychological stress level detection based on heartbeat mode |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |