CN109994216A - A kind of ICD intelligent diagnostics coding method based on machine learning - Google Patents
A kind of ICD intelligent diagnostics coding method based on machine learning Download PDFInfo
- Publication number
- CN109994216A CN109994216A CN201910218254.3A CN201910218254A CN109994216A CN 109994216 A CN109994216 A CN 109994216A CN 201910218254 A CN201910218254 A CN 201910218254A CN 109994216 A CN109994216 A CN 109994216A
- Authority
- CN
- China
- Prior art keywords
- icd
- diagnosis
- coding
- machine learning
- preliminary classification
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- 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
- G16H70/00—ICT specially adapted for the handling or processing of medical references
Abstract
The present invention provides a kind of ICD intelligent diagnostics coding method based on machine learning, comprising the following steps: access data center extracts diagnosis and treatment data and artificial ICD diagnoses code;Diagnosis and treatment data are converted to numeric type feature vector by the text feature for extracting diagnosis and treatment data;To the sample data stochastical sampling and feature selecting of feature vector composition, different preliminary classification devices is constructed using different machine learning methods, the classifier for recycling artificial ICD diagnosis code training different is until convergence;The input of output result composition Secondary classifier after the convergence of preliminary classification device, further according to artificial ICD diagnosis code training Secondary classifier until convergence;When new, uncoded diagnosis and treatment data input, ICD diagnosis coding is exported according to feature extraction and trained primary, Secondary classifier, computer automatically.The method increase code efficiency and coding quality, hospital clinical path code standardized to ICD and DRGs payment management are using significant.
Description
Technical field
The present invention relates to biomedical and information technology field, more specifically, is related to a kind of based on machine learning
ICD intelligent diagnostics coded system.
Background technique
International Classification of Diseases (ICD) is the basis for determining global health trends and statistical data, wherein containing about 55000
Unique code related with damage, disease and the cause of the death has morbidity statistics, hospital management, medical treatment payment and medical research etc.
Significance.It helps to collect and store medical data in order to analyze and carry out decision in evidence-based, makes the whole world
It can be compared between hospital, region and country and between different times and sharing data in a manner of consistent and standard.Together
When, China, which is being widelyd popularize, is grouped (DRGs) pre-payment system by disease dependent diagnostic.DRGs is comparison generally acknowledged in the world
One of advanced means of payment, according to the age of patient, gender, length of stay, clinical diagnosis, illness, operation, the serious journey of disease
Degree, complication and complication and the factors such as lapses to patient is divided into 500-600 diagnosis relevant group, the progress science survey in grouping
It calculates, gives imprest money.And ICD coding is then the standardization to diagnosing patient, is that DRGs Payment system can play a role
Basis and China's medical treatment control expense successful implementation guarantee.
The diagnosis coding work of hospital, China is mainly write by coder with reference to doctor at present medical document is completed,
This not only expends the time, but also is easy error.Firstly, medical document is write by doctor, everyone communicative habits are not
Together, using different abbreviations and synonym, it is easy to cause the ambiguousness of coding, secondly as contacting between the disease of part tight
It is close, need to be combined together comprehensive coding, some unfamiliar coders would generally be separately encoded by disease, in addition, ICD
Coding is the coded system of a level, is divided into chapter, section, classification, suborder etc., and coder often distributes one to disease description
Excessively wide in range coding leads to the inaccurate of morbidity statistics.With Hospital Electronic Medical Record application deeply, some hospitals are by ICD
Coding work gives clinician to complete, but the effect is unsatisfactory.
In order to promote code efficiency, improve coding quality, a series of Computer-Aided Coding technology is all suggested and reality
It is existing.Firstly, common ICD coded system mainly encodes retrieval service on the market at present, i.e., preferred building encoding name dictionary,
Including disease name and corresponding disease code, after coder inputs the diagnosis description in electronic health record, system is pressed
It is retrieved from encoding name dictionary according to title, returns to candidate code and selected for coder.But many doctors are due to region, environment
The diagnosis description of equal communicative habits problem, writing is often lack of standardization and cannot understand mutually, it is thus possible to which retrieval is less than corresponding
ICD coding.And each clinician has its common clinical diagnosis, if coder every doctor is established respectively it is respective
Common diagnosis dictionary corresponding with ICD coding, then need to expend great effort and clinician's communication, and is difficult exhaustive all face
Bed diagnosis.Secondly, with the development of artificial intelligence technology, gradually appeared using natural language processing technique and machine learning into
The mode of row ICD coding.Natural language processing technique mainly carries out Chinese word segmentation to the clinical diagnosis of electronic health record, then again and
Title carries out measuring similarity, obtains candidate code according to similarity distance.But the usual sentence of clinical diagnosis is shorter, point
Word result is similar with direct matching result, and it is single obtain ICD coding from clinical diagnosis, may due to doctor's clerical error and
Encoding error is caused, the coding quality of ICD cannot be effectively improved.There are also application of the machine learning techniques on ICD intelligently encoding.
Machine learning mainly passes through extraction feature, then utilizes feature train classification models, prediction ICD coding.But common machines
What habit technology only encoded corresponding relationship from electronic health record to ICD assumes that study in space is optimal to one it is assumed that different study
Model frequently results in different corresponding relationships, on whole prediction effect and not bery satisfactory.
Given this kind of area of computer aided ICD intelligent diagnostics coding method of efficiently and accurately is needed, to improve ICD coding matter
Amount and code efficiency, medical worker is freed from heavy, duplicate work, is established for China's hospital management and medical treatment control expense
Determine basis for IT application.
Summary of the invention
The ICD intelligent diagnostics coding method based on machine learning that the purpose of the present invention is a kind of is used for electronic health record certainly
It is dynamic to assign ICD coding, to improve coding quality and code efficiency, establish the basis of hospital informatioization management.It is above-mentioned in order to realize
Purpose, the present invention the following technical schemes are provided:
A kind of ICD intelligent diagnostics coding method based on machine learning, comprising the following steps:
S1: data extraction module accesses the data center of medical institutions, and patient-centered health care extracts relevant diagnosis and treatment data
Code is diagnosed with the ICD of h coding;
S2: text feature is extracted to the diagnosis and treatment data being drawn into using characteristic extracting module, diagnosis and treatment data are converted into number
The feature vector of value type;
S3: to the sample data stochastical sampling and feature selecting of described eigenvector composition, different machines is respectively adopted
Learning method constructs different preliminary classification device modules, using the different classifier of corresponding artificial ICD diagnosis code training until
Convergence keeps classifier diversified as far as possible each other on the basis of respectively meeting preferable accuracy;S4: preliminary classification device module convergence
The input of output result composition Secondary classifier module afterwards, again according to corresponding artificial ICD diagnosis code training subclassificatio
Device module is until convergence;
S5: when new, uncoded diagnosis and treatment data input, according to characteristic extracting module and trained preliminary classification
Device module and Secondary classifier module, computer export ICD diagnosis coding automatically.
Further, the data extraction module includes the data acquisition interface to medical institutions.
Further, step S2 includes the following contents:
Chinese word segmentation: collected diagnosis and treatment data are tentatively segmented by maximum matching method, are built using statistical method
Dedicated medical treatment dictionary, to preliminary word segmentation result disambiguation;
Bag of words: keyword is extracted with bag of words, the weight of keyword is indicated with word frequency inverse document frequency;
Topic model: the theme for representing text is extracted using the distribution of implicit Di Li Cray.
Further, step S3 includes the following contents:
Over-sampling: sample size inconsistence problems caused by the frequency difference occurred for disease, using synthesis minority class
Oversampling technique balance sample size;
Feature selecting: different features is selected to different preliminary classification devices, decision tree increases the demography letter of patient
Breath, obtains the mapping relations of patient population information and ICD coding at gender, age etc., and naive Bayesian then has ignored feelings of being admitted to hospital
The theme feature of condition construction, increases the diversity between preliminary classification device;
Learning object: different preliminary classification devices, including support vector machines, decision tree, naive Bayesian, k nearest neighbor are used
And neural network.
Further, step S4 includes the following contents:
The output that different preliminary classification devices are combined by the combination strategy of learning method is further learned using support vector machines
Practise the mapping relations of different preliminary classification devices and final prediction result.
The utility model has the advantages that ICD intelligent diagnostics coding method provided by the invention can be in existing patient's diagnosis and treatment data basis
On, establish machine learning model;When coder carries out ICD coding to new clinical diagnosis, the diagnosis and treatment data of patient are inputted,
This method carries out Chinese word segmentation, feature extraction automatically, then by preliminary classification device and Secondary classifier, can intelligently export
Most probable coding.The coding method of ICD intelligent diagnostics can learn doctor's writing using a large amount of h coding's data set
Mapping relations between medical document and ICD coding, while can constantly be mentioned with the increase of data volume, increasing for validity feature
The accuracy of height prediction, helps coder efficiently and accurately to complete ICD coding work.This is in the application of clinical electronic health record
ICD is code standardized, hospital clinical path and DRGs payment management are using significant.
Detailed description of the invention
Fig. 1 is the processing flow schematic diagram of ICD intelligent diagnostics coding method of the present invention.
Fig. 2 is the feature selection step schematic diagram of ICD intelligent diagnostics coding method of the present invention.
Fig. 3 is the preliminary classification device step schematic diagram of ICD intelligent diagnostics coding method of the present invention.
Fig. 4 is the Secondary classifier step schematic diagram of ICD intelligent diagnostics coding method of the present invention.
Specific embodiment
The invention will now be further described with reference to specific embodiments, but examples are merely exemplary, not to this hair
Bright range constitutes any restrictions.It will be understood by those skilled in the art that without departing from the spirit and scope of the invention
Can with the details and forms of the technical scheme of the invention are modified or replaced, but these modification and replacement each fall within it is of the invention
In protection scope.
A specific embodiment of the invention is described in detail below with reference to attached drawing.
The present invention provides a kind of ICD intelligent diagnostics coding method based on machine learning, comprising the following steps:
S1: data extraction module accesses the data center of medical institutions, and patient-centered health care extracts relevant diagnosis and treatment data
Code is diagnosed with the ICD of h coding;
S2: text feature is extracted to the diagnosis and treatment data being drawn into using characteristic extracting module, diagnosis and treatment data are converted into number
The feature vector of value type;
S3: to the sample data stochastical sampling and feature selecting of described eigenvector composition, different machines is respectively adopted
Learning method constructs different preliminary classification device modules, using the different classifier of corresponding artificial ICD diagnosis code training until
Convergence keeps classifier diversified as far as possible each other on the basis of respectively meeting preferable accuracy;
S4: the input of the output result composition Secondary classifier module after the convergence of preliminary classification device module, again according to phase
The artificial ICD diagnosis code training Secondary classifier module answered is until convergence;
S5: when new, uncoded diagnosis and treatment data input, according to characteristic extracting module and trained preliminary classification
Device module and Secondary classifier module, computer export ICD diagnosis coding automatically.
Specifically, ICD intelligent diagnostics coded system provided by the invention includes following module:
1. data extraction module
Data extraction module contains the data acquisition interface to medical institutions, the seamless hospital information system with hospital
(HIS), radiology information system (RIS), image filing and communication system (PACS), inspection checking system (LIS), nurse's information system
The docking of the information systems such as system (NIS), clinic information system (CIS) and electronic medical record system (EMR), patient-centered health care acquire phase
The diagnosis and treatment data of pass, including situation of being admitted to hospital, inspection inspection result, patient diagnosis, admission diagnosis, discharge diagnosis and patient are basic
Information etc..
2. characteristic extracting module
The text data that characteristic extracting module is obtained mainly for data extraction module carries out feature extraction, utilizes nature language
The medical document that doctor writes is converted feature vector by the method for speech processing.Bag of words are that natural language processing kind is the most frequently used
One of model, all texts are regarded as a series of set of words by it, and ignore the sequence of word.Particularly as being by whole section of text
It being separated as unit of word, every article can be expressed as a long vector, a word is represented per one-dimensional in vector, and the dimension
Corresponding weight represents significance level of this word in article, and usually we use word frequency inverse document frequency (Term
Frequency-Inverse Document Frequency, TF-IDF) indicate the weight of word.Medical text includes doctor couple
The word habit of the description of symptom, the cause of disease when patient is admitted to hospital etc., these description contents varies with each individual, and there are a large amount of synonyms
And the phenomenon that polysemy, TF-IDF cannot effectively extract text feature at this time, and therefore, we use topic model to retouch
It states.According to above-mentioned introduction, medical text is represented as the feature vector characterized by the sum that TF-IDF is characterized by topic model, each
The diagnosis and treatment data of patient are all converted into a feature vector.
3. preliminary classification device module
Integrated study is a kind of machine learning method solved the same problem by the multiple classifiers of training, compared to more general
Logical machine learning method learns for a hypothesis from training set, and a series of hypothesis of integrated study trial learning is simultaneously incorporated into
Get up to use.Integrated study mentions high performance on condition that each preliminary classification device " good and different ", i.e. preliminary classification device are guaranteeing standard
On the basis of true property, there is diversity as far as possible each other.Heterogeneous integrated study uses different types of classifier, itself just has
There is diversity.Therefore the method that present example uses heterogeneous integrated study, and primary is further increased by attribute disturbance
The diversity of classifier.
4. Secondary classifier module
Secondary classifier is a kind of combination preliminary classification device result and then generates the better classifier of overall effect, is usually adopted
Final result is provided with ballot method, but the feature of preliminary classification device cannot be fully utilized in method of voting.Learning strategy is a kind of
The method that multiple heterogeneous classifiers can be combined, preliminary classification device no longer directly exports the classification of prediction at this time, but exports and belong to
In the probability distribution of each classification.The input that the present embodiment connects the output vector of preliminary classification device as Secondary classifier,
The weight of each preliminary classification device probability distribution is finally obtained, ICD coding result is exported.
Above with reference to attached drawing, a kind of ICD intelligent diagnostics coding method based on machine learning proposed according to the present invention is retouched
One embodiment is stated.It will be understood by those skilled in the art, however, that the ICD intelligent diagnostics proposed for aforementioned present invention
Coded system can also make various improvement on the basis of not departing from the content of present invention.Therefore, protection scope of the present invention is answered
It is determined when by the content of appended claims.
Claims (5)
1. a kind of ICD intelligent diagnostics coding method based on machine learning, which comprises the following steps:
S1: data extraction module accesses the data center of medical institutions, and patient-centered health care extracts relevant diagnosis and treatment data and people
The ICD of work coding diagnoses code;
S2: text feature is extracted to the diagnosis and treatment data being drawn into using characteristic extracting module, diagnosis and treatment data are converted into numeric type
Feature vector;
S3: to the sample data stochastical sampling and feature selecting of described eigenvector composition, different machine learning is respectively adopted
The different preliminary classification device module of method construct, using the different classifier of corresponding artificial ICD diagnosis code training until restraining,
Keep classifier diversified as far as possible each other on the basis of respectively meeting preferable accuracy;
S4: the input of the output result composition Secondary classifier module after the convergence of preliminary classification device module, again according to corresponding
Artificial ICD diagnosis code training Secondary classifier module is until convergence;
S5: when new, uncoded diagnosis and treatment data input, according to characteristic extracting module and trained preliminary classification device mould
Block and Secondary classifier module, computer export ICD diagnosis coding automatically.
2. a kind of ICD intelligent diagnostics coding method based on machine learning according to claim 1, which is characterized in that institute
Stating data extraction module includes the data acquisition interface to medical institutions.
3. a kind of ICD intelligent diagnostics coding method based on machine learning according to claim 1, which is characterized in that step
Rapid S2 includes the following contents:
Chinese word segmentation: collected diagnosis and treatment data are tentatively segmented by maximum matching method, are built using statistical method dedicated
Medical dictionary, to preliminary word segmentation result disambiguation;
Bag of words: keyword is extracted with bag of words, the weight of keyword is indicated with word frequency inverse document frequency;
Topic model: the theme for representing text is extracted using the distribution of implicit Di Li Cray.
4. a kind of ICD intelligent diagnostics coding method based on machine learning according to claim 3, which is characterized in that step
Rapid S3 includes the following contents:
Over-sampling: sample size inconsistence problems caused by the frequency difference occurred for disease are crossed using synthesis minority class and are adopted
Sample technology balance sample size;
Feature selecting: different features is selected to different preliminary classification devices, decision tree increases the demographic of patient, property
Not, age etc., the mapping relations of patient population information and ICD coding are obtained, naive Bayesian then has ignored situation structure of being admitted to hospital
The theme feature made increases the diversity between preliminary classification device;
Learning object: different preliminary classification devices, including support vector machines, decision tree, naive Bayesian, k nearest neighbor and mind are used
Through network.
5. a kind of ICD intelligent diagnostics coding method based on machine learning according to claim 4, which is characterized in that step
Rapid S4 includes the following contents:
The output that different preliminary classification devices are combined by the combination strategy of learning method is further learnt not using support vector machines
With the mapping relations of preliminary classification device and final prediction result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910218254.3A CN109994216A (en) | 2019-03-21 | 2019-03-21 | A kind of ICD intelligent diagnostics coding method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910218254.3A CN109994216A (en) | 2019-03-21 | 2019-03-21 | A kind of ICD intelligent diagnostics coding method based on machine learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109994216A true CN109994216A (en) | 2019-07-09 |
Family
ID=67129093
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910218254.3A Pending CN109994216A (en) | 2019-03-21 | 2019-03-21 | A kind of ICD intelligent diagnostics coding method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109994216A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110491465A (en) * | 2019-08-20 | 2019-11-22 | 山东众阳健康科技集团有限公司 | Classification of diseases coding method, system, equipment and medium based on deep learning |
CN110837494A (en) * | 2019-10-12 | 2020-02-25 | 云知声智能科技股份有限公司 | Method and device for identifying unspecified diagnosis coding errors of medical record home page |
CN110895580A (en) * | 2019-12-12 | 2020-03-20 | 山东众阳健康科技集团有限公司 | ICD operation and operation code automatic matching method based on deep learning |
CN110970103A (en) * | 2019-10-09 | 2020-04-07 | 北京雅丁信息技术有限公司 | Method for searching correlation between diagnosis and medicine in electronic medical record |
CN110991170A (en) * | 2019-12-05 | 2020-04-10 | 清华大学 | Chinese disease name intelligent standardization method and system based on electronic medical record information |
CN111402974A (en) * | 2020-03-06 | 2020-07-10 | 西南交通大学 | Electronic medical record ICD automatic coding method based on deep learning |
CN111462896A (en) * | 2020-03-31 | 2020-07-28 | 重庆大学 | Real-time intelligent auxiliary ICD coding system and method based on medical record |
CN111899829A (en) * | 2020-07-31 | 2020-11-06 | 青岛百洋智能科技股份有限公司 | Full-text retrieval matching engine based on ICD9/10 participle lexicon |
CN112530582A (en) * | 2020-12-11 | 2021-03-19 | 万达信息股份有限公司 | Intelligent system for assisting cause of death classified coding |
CN112599213A (en) * | 2021-03-04 | 2021-04-02 | 联仁健康医疗大数据科技股份有限公司 | Classification code determining method, device, equipment and storage medium |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005107603A (en) * | 2003-09-26 | 2005-04-21 | Yuyama Manufacturing Co Ltd | Icd code applying device |
US20080288292A1 (en) * | 2007-05-15 | 2008-11-20 | Siemens Medical Solutions Usa, Inc. | System and Method for Large Scale Code Classification for Medical Patient Records |
CN104504366A (en) * | 2014-11-24 | 2015-04-08 | 上海闻泰电子科技有限公司 | System and method for smiling face recognition based on optical flow features |
US20150379241A1 (en) * | 2014-06-27 | 2015-12-31 | Passport Health Communications, Inc. | Automatic medical coding system and method |
CN106844308A (en) * | 2017-01-20 | 2017-06-13 | 天津艾登科技有限公司 | A kind of use semantics recognition carries out the method for automating disease code conversion |
CN107766792A (en) * | 2017-06-23 | 2018-03-06 | 北京理工大学 | A kind of remote sensing images ship seakeeping method |
CN109065157A (en) * | 2018-08-01 | 2018-12-21 | 中国人民解放军第二军医大学 | A kind of Disease Diagnosis Standard coded Recommendation list determines method and system |
CN109145108A (en) * | 2017-06-16 | 2019-01-04 | 贵州小爱机器人科技有限公司 | Classifier training method, classification method, device and computer equipment is laminated in text |
CN109273062A (en) * | 2018-08-09 | 2019-01-25 | 北京爱医声科技有限公司 | ICD intelligence Auxiliary Encoder System |
CN109299741A (en) * | 2018-06-15 | 2019-02-01 | 北京理工大学 | A kind of network attack kind identification method based on multilayer detection |
CN109324604A (en) * | 2018-11-29 | 2019-02-12 | 中南大学 | A kind of intelligent train resultant fault analysis method based on source signal |
-
2019
- 2019-03-21 CN CN201910218254.3A patent/CN109994216A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005107603A (en) * | 2003-09-26 | 2005-04-21 | Yuyama Manufacturing Co Ltd | Icd code applying device |
US20080288292A1 (en) * | 2007-05-15 | 2008-11-20 | Siemens Medical Solutions Usa, Inc. | System and Method for Large Scale Code Classification for Medical Patient Records |
US20150379241A1 (en) * | 2014-06-27 | 2015-12-31 | Passport Health Communications, Inc. | Automatic medical coding system and method |
CN104504366A (en) * | 2014-11-24 | 2015-04-08 | 上海闻泰电子科技有限公司 | System and method for smiling face recognition based on optical flow features |
CN106844308A (en) * | 2017-01-20 | 2017-06-13 | 天津艾登科技有限公司 | A kind of use semantics recognition carries out the method for automating disease code conversion |
CN109145108A (en) * | 2017-06-16 | 2019-01-04 | 贵州小爱机器人科技有限公司 | Classifier training method, classification method, device and computer equipment is laminated in text |
CN107766792A (en) * | 2017-06-23 | 2018-03-06 | 北京理工大学 | A kind of remote sensing images ship seakeeping method |
CN109299741A (en) * | 2018-06-15 | 2019-02-01 | 北京理工大学 | A kind of network attack kind identification method based on multilayer detection |
CN109065157A (en) * | 2018-08-01 | 2018-12-21 | 中国人民解放军第二军医大学 | A kind of Disease Diagnosis Standard coded Recommendation list determines method and system |
CN109273062A (en) * | 2018-08-09 | 2019-01-25 | 北京爱医声科技有限公司 | ICD intelligence Auxiliary Encoder System |
CN109324604A (en) * | 2018-11-29 | 2019-02-12 | 中南大学 | A kind of intelligent train resultant fault analysis method based on source signal |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110491465A (en) * | 2019-08-20 | 2019-11-22 | 山东众阳健康科技集团有限公司 | Classification of diseases coding method, system, equipment and medium based on deep learning |
WO2021032219A3 (en) * | 2019-08-20 | 2021-04-15 | 山东众阳健康科技集团有限公司 | Method and system for disease classification coding based on deep learning, and device and medium |
CN110970103A (en) * | 2019-10-09 | 2020-04-07 | 北京雅丁信息技术有限公司 | Method for searching correlation between diagnosis and medicine in electronic medical record |
CN110837494A (en) * | 2019-10-12 | 2020-02-25 | 云知声智能科技股份有限公司 | Method and device for identifying unspecified diagnosis coding errors of medical record home page |
CN110837494B (en) * | 2019-10-12 | 2022-03-25 | 云知声智能科技股份有限公司 | Method and device for identifying unspecified diagnosis coding errors of medical record home page |
CN110991170B (en) * | 2019-12-05 | 2021-10-12 | 清华大学 | Chinese disease name intelligent standardization method and system based on electronic medical record information |
CN110991170A (en) * | 2019-12-05 | 2020-04-10 | 清华大学 | Chinese disease name intelligent standardization method and system based on electronic medical record information |
CN110895580A (en) * | 2019-12-12 | 2020-03-20 | 山东众阳健康科技集团有限公司 | ICD operation and operation code automatic matching method based on deep learning |
CN110895580B (en) * | 2019-12-12 | 2020-07-07 | 山东众阳健康科技集团有限公司 | ICD operation and operation code automatic matching method based on deep learning |
CN111402974A (en) * | 2020-03-06 | 2020-07-10 | 西南交通大学 | Electronic medical record ICD automatic coding method based on deep learning |
CN111462896A (en) * | 2020-03-31 | 2020-07-28 | 重庆大学 | Real-time intelligent auxiliary ICD coding system and method based on medical record |
CN111462896B (en) * | 2020-03-31 | 2023-04-18 | 重庆大学 | Real-time intelligent auxiliary ICD coding system and method based on medical record |
CN111899829A (en) * | 2020-07-31 | 2020-11-06 | 青岛百洋智能科技股份有限公司 | Full-text retrieval matching engine based on ICD9/10 participle lexicon |
CN112530582A (en) * | 2020-12-11 | 2021-03-19 | 万达信息股份有限公司 | Intelligent system for assisting cause of death classified coding |
CN112530582B (en) * | 2020-12-11 | 2023-11-14 | 万达信息股份有限公司 | Intelligent system for assisting classified coding of death cause |
CN112599213B (en) * | 2021-03-04 | 2021-05-25 | 联仁健康医疗大数据科技股份有限公司 | Classification code determining method, device, equipment and storage medium |
CN112599213A (en) * | 2021-03-04 | 2021-04-02 | 联仁健康医疗大数据科技股份有限公司 | Classification code determining method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109994216A (en) | A kind of ICD intelligent diagnostics coding method based on machine learning | |
CN111540468B (en) | ICD automatic coding method and system for visualizing diagnostic reasons | |
WO2022227294A1 (en) | Disease risk prediction method and system based on multi-modal fusion | |
CN112214995A (en) | Hierarchical multitask term embedding learning for synonym prediction | |
WO2019132685A1 (en) | Method and system for supporting medical decision making | |
CN109378066A (en) | A kind of control method and control device for realizing disease forecasting based on feature vector | |
López-Úbeda et al. | Automatic medical protocol classification using machine learning approaches | |
CN111949759A (en) | Method and system for retrieving medical record text similarity and computer equipment | |
CN110364234A (en) | Electronic health record intelligent storage analyzing search system and method | |
Wang et al. | A framework and its empirical study of automatic diagnosis of traditional Chinese medicine utilizing raw free-text clinical records | |
CN112489740A (en) | Medical record detection method, training method of related model, related equipment and device | |
Kaswan et al. | AI-based natural language processing for the generation of meaningful information electronic health record (EHR) data | |
CN111785387A (en) | Method and system for disease standardized mapping classification by using Bert | |
Chen et al. | A deep-learning based ultrasound text classifier for predicting benign and malignant thyroid nodules | |
CN113555077A (en) | Suspected infectious disease prediction method and device | |
CN110164519B (en) | Classification method for processing electronic medical record mixed data based on crowd-sourcing network | |
Zhou et al. | Tamfn: Time-aware attention multimodal fusion network for depression detection | |
WO2019132686A1 (en) | Method for generating mathematical models of a patient using artificial intelligence technologies | |
Waheeb et al. | An efficient sentiment analysis based deep learning classification model to evaluate treatment quality | |
CN116881336A (en) | Efficient multi-mode contrast depth hash retrieval method for medical big data | |
Zhang et al. | BiomedCLIP: a multimodal biomedical foundation model pretrained from fifteen million scientific image-text pairs | |
Zhang et al. | Deep holistic representation learning from ehr | |
US11809826B2 (en) | Assertion detection in multi-labelled clinical text using scope localization | |
Alsaidi et al. | Exploring analogical inference in healthcare | |
Wang et al. | Automatic ICD-10 coding based on multi-head attention mechanism and gated residual network |
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 |