CN115662647B - Method for excavating similar diseases and application - Google Patents
Method for excavating similar diseases and application Download PDFInfo
- Publication number
- CN115662647B CN115662647B CN202211679383.0A CN202211679383A CN115662647B CN 115662647 B CN115662647 B CN 115662647B CN 202211679383 A CN202211679383 A CN 202211679383A CN 115662647 B CN115662647 B CN 115662647B
- Authority
- CN
- China
- Prior art keywords
- disease
- diseases
- knowledge graph
- medicine
- vector representation
- 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
- 201000010099 disease Diseases 0.000 title claims abstract description 169
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 title claims abstract description 169
- 238000000034 method Methods 0.000 title claims abstract description 24
- 239000013598 vector Substances 0.000 claims abstract description 58
- 238000005295 random walk Methods 0.000 claims abstract description 13
- 239000003814 drug Substances 0.000 claims description 49
- 208000024891 symptom Diseases 0.000 claims description 23
- 229940079593 drug Drugs 0.000 claims description 22
- 238000005065 mining Methods 0.000 claims description 20
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000003745 diagnosis Methods 0.000 claims description 3
- 206010034464 Periarthritis Diseases 0.000 description 5
- 206010028836 Neck pain Diseases 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 230000001225 therapeutic effect Effects 0.000 description 2
- 229940126680 traditional chinese medicines Drugs 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 208000019693 Lung disease Diseases 0.000 description 1
- 208000022873 Ocular disease Diseases 0.000 description 1
- 238000012550 audit Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000008187 granular material Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- 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
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention relates to a method for excavating similar diseases and application thereof, wherein the method comprises the following steps: step one, constructing a knowledge graph: the knowledge graph comprises a pharmaceutical knowledge graph and a disease knowledge graph; step two, obtaining vector representation of the disease based on the knowledge graph: in order to obtain vector representation of the disease, a plurality of node sequences are obtained in a random walk mode; taking the disease as an initial node, acquiring a next-hop node according to the connection relation among the nodes, and the like; generating a vector representation of the disease by a Word2Vec model after the plurality of node sequences are acquired; step three, evaluating similarity based on vector representation of the disease: and calculating cosine similarity of the two diseases according to vector representation of the diseases, wherein the numerical value of the cosine similarity is the similarity between the two diseases.
Description
Technical Field
The invention relates to the technical field of intelligent medical treatment, in particular to a method for excavating similar diseases and application thereof.
Background
The method for mining similar diseases can be used as a part of a prescription auditing system, and is mainly used for mining the correlation among different diagnosis terms, so as to provide reference for a pharmacist to audit an electronic prescription.
The main method for evaluating similarity between diseases in the prior art is also based on editing distance. For example, chinese patent publication No. CN105095665B discloses a method for structuring disease information, which resolves a disease into information such as a disease occurrence part, a disease degree, a disease body, etc. according to a preset dimension, and compares the information of the two diseases to achieve the purpose of evaluating the disease similarity.
Because of the variety of descriptions of diseases, a great deviation exists in evaluating similarity among diseases based on the edit distance alone, which directly leads to a high false positive rate of disease matching.
Disclosure of Invention
The invention aims to provide a method for mining similar diseases, which aims to solve the problem of how to improve the accuracy of similarity measurement between diseases.
The invention aims to solve the defects of the prior art and provides a method for excavating similar diseases, which comprises the following steps:
step one, constructing a knowledge graph: the knowledge graph comprises a pharmaceutical knowledge graph and a disease knowledge graph;
step two, obtaining vector representation of the disease based on the knowledge graph: in order to obtain vector representation of the disease, a plurality of node sequences are obtained in a random walk mode; taking the disease as an initial node, acquiring a next-hop node according to the connection relation among the nodes, and the like; generating a vector representation of the disease by a Word2Vec model after the plurality of node sequences are acquired;
step three, evaluating similarity based on vector representation of the disease: and calculating cosine similarity of the two diseases according to vector representation of the diseases, wherein the numerical value of the cosine similarity is the similarity between the two diseases.
Preferably, the pharmaceutical knowledge graph comprises main components of medicines, treatment sites, treatment disease information, ATC coding and ICD10 coding information; the disease knowledge graph comprises treatment of diseases, disease parts, common symptoms, ICD10 codes and treatment relations between medicines and diseases; the main components, the treatment parts and the treatment disease information of the medicine are obtained through the instruction book of the medicine; the ICD10 coding information is supplemented by ICD10 coding for treating diseases; the ATC code is obtained by reasoning main components of the medicine; treatment of disease, site of onset, common symptoms, ICD10 codes are extracted from disease encyclopedia; the therapeutic relationship of drugs to diseases is obtained through a plurality of electronic prescriptions.
Preferably, the vector representation of the disease generated by the Word2Vec model is divided into two steps of node sequence generation and disease vector calculation.
Preferably, the generation of the node sequence specifically means that the node sequence is constructed by random walk on a knowledge graph.
Preferably, the random walk on the knowledge graph specifically refers to random walk according to a meta-path.
Preferably, the meta-path refers to a medically interpretable path.
Preferably, the meta-path specifically includes 1) drug-disease; 2) Drug-symptom-disease; 3) Medicine-principal component-disease; 4) Medicine-site-disease; 5) drug-ICD 10 code-disease.
Preferably, nodes sampled by different meta-paths can be spliced by the same node to generate a longer sequence, and more nodes of different drugs or diseases can be spliced into one sequence by disease nodes or drug nodes.
Preferably, in the disease vector calculation, the input of the disease vector is a plurality of node sequences generated based on the migration of the knowledge graph; the node sequence may include a plurality of disease nodes associated with each other by drug nodes and symptom nodes, i.e., the associated disease nodes are more likely to be treated by similar drugs or include the same similar symptoms, i.e., have similar medical manifestations; generating vector representation of each node in the knowledge graph by adopting a universal Word2Vec algorithm; the disease vector generated in the way can ensure that the cosine similarity of vector representation of the disease with more similar medical performance is larger.
Preferably, the vector representation of the high-dimensional disease is reduced to a 2-dimensional vector using the t-SNE algorithm and plotted on a 2D coordinate system.
The invention also provides application of the method for mining similar diseases in mining the relationship between the traditional Chinese medicines and/or expanding the relationship between the traditional Chinese medicines in the prescription of the multiple medicines.
Advantageous effects
Compared with the prior art, the invention has the beneficial effects that:
the method for mining similar diseases is a similarity measurement method based on a knowledge graph, and the core assumption is that the more the two diseases have the same symptoms, the more similar the two diseases are; it can be further assumed that if two diseases have more identical neighboring nodes on the knowledge graph, the more similar the two diseases are.
The method for mining similar diseases adopts a universal Word2Vec algorithm to generate vector representation of each node in the knowledge graph. The disease vector generated in this way can ensure that the more similar the medical manifestation is, the greater the cosine similarity of the vector representation of the disease. Meanwhile, the invention adopts the t-SNE algorithm to reduce the dimension of the high-dimension vector of the disease to the 2-dimension vector, and draws the vector on a 2D coordinate system, so that the distance between two diseases can be represented on a 2D plan view.
The method for mining similar diseases is mainly applied to two scenes:
1. the relation of the traditional Chinese medicine symptoms in the prescription of multiple medicines is excavated. The keyword matching mode cannot determine which kind of diseases in the prescription is treated by each medicine, and based on the invention, the most similar diagnosis of the indications of medicine treatment and the prescription can be evaluated, so that the accuracy of medicine relation mining is improved.
2. Expanding the relationship of medicine symptoms. The invention can realize the clustering of diseases, and if a certain medicine can treat most diseases in a certain category, the medicine can treat other diseases in the category with high probability, so that the medicine disease relation can be expanded.
Both the above two scenes are effective to the pharmacist, and the final extraction result requires the pharmacist to review.
The method for mining similar diseases can generate vector representation of the diseases based on a random walk algorithm of a knowledge graph, and evaluate the similarity of the diseases through cosine similarity, and the effect is obviously better than that of a method based on editing distance.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate and do not limit the invention.
Fig. 1 is a diagram of the structure of a pharmaceutical knowledge graph and a disease knowledge graph.
Fig. 2 is a schematic representation of a sample of the results of mining for similar diseases.
Fig. 3 is a knowledge-graph sample illustration.
Detailed Description
The present invention is described in more detail below to facilitate an understanding of the present invention.
The complete structures of the pharmaceutical knowledge graph and the disease knowledge graph used in the invention are shown in figure 1.
The method for excavating similar diseases comprises the following steps:
step one, constructing a knowledge graph shown in fig. 1: the knowledge graph comprises a pharmaceutical knowledge graph and a disease knowledge graph;
step two, obtaining vector representation of the disease based on the knowledge graph: in order to obtain vector representation of the disease, the invention adopts a random walk mode to obtain a large number of node sequences; and taking the disease as an initial node, acquiring a next-hop node according to the connection relation among the nodes, and the like. After a large number of node sequences are acquired, the present invention generates a vector representation of the disease through the Word2Vec model.
The Word2Vec algorithm is a group of correlation models used to generate Word vectors. These models are shallow, bi-layer neural networks that are used to train to reconstruct linguistic word text. The network is represented by words and guesses the input words in adjacent positions, and the order of the words is unimportant under the word bag model assumption in word2 vec. After training is completed, word2vec models can be used to map each word to a vector that can be used to represent word-to-word relationships, which is the hidden layer of the neural network.
Specifically, generating a vector representation of a disease by Word2Vec model is largely divided into two steps: and generating a node sequence and calculating a disease vector.
1) The node sequence is mainly constructed by random walk on the knowledge graph; however, to introduce interpretability, the present invention introduces the concept of a meta path, which is used by the present invention to include: 1) Drug-disease; 2) Drug-symptom-disease; 3) Medicine-principal component-disease; 4) Medicine-site-disease; 5) drug-ICD 10 code-disease. The introduction of the meta-path concept ensures that the node sequences randomly generated by the method have medical rationality, thereby having interpretability. The present invention is not limited to the above-described five-element path, and any path that is medically interpretable may be used as the element path.
Taking fig. 1 as an example, it is further explained why the present invention has medical interpretability. The invention generates a node sequence on the knowledge graph shown in figure 3; the element path 4) can be used for sampling 'Gentongping particles-shoulder-left scapulohumeral periarthritis', and the path can be interpreted as that the Gentongping particles can treat the shoulder of the part, and the disease part of the left scapulohumeral periarthritis is the shoulder; the element path 2) can be sampled to obtain the 'Gentongping particles-shoulder neck pain-left scapulohumeral periarthritis', and the path can be interpreted as that the Gentongping particles can treat shoulder neck pain, and the left scapulohumeral periarthritis has a typical symptom of shoulder neck pain.
Furthermore, the nodes sampled by the meta-path 2) and the meta-path 4) can be spliced through the same node 'Gentongping granule' and 'left scapulohumeral periarthritis' to generate a longer sequence, and more nodes with different medicines or diseases can be spliced into a sequence through disease nodes or medicine nodes.
2) The disease vector is input as a plurality of node sequences generated based on the migration of the knowledge-based maps. As previously mentioned, these sequences may include a large number of disease nodes associated by nodes such as drug nodes, symptom nodes, etc., i.e., these associated disease nodes are more likely to be treated by similar drugs or include the same similar symptoms, i.e., have similar medical manifestations.
The invention adopts a universal Word2Vec algorithm to generate vector representation of each node in the knowledge graph. The disease vector generated in this way can ensure that the more similar the medical manifestation is, the greater the cosine similarity of the vector representation of the disease.
And thirdly, evaluating the similarity based on vector representation of the disease. In order to evaluate the similarity of two diseases, the invention firstly obtains vector expression of the diseases, and calculates cosine similarity of the two diseases according to the vector expression, wherein the value is the similarity between the diseases.
The vector representation of the disease is a high-dimensional vector representation, and it is difficult to represent the distance between two diseases in a single 2D planar figure. Therefore, the invention adopts the t-SNE algorithm to reduce the dimension of the high-dimension vector of the disease to the 2-dimension vector, and draws the vector on a 2D coordinate system. Fig. 2 shows a sample of similarity mining results generated by the present invention, it can be seen that more of the upper left hand corner is ocular disease, the lower left hand corner is pulmonary disease, and more of the right hand side is dermatological disease. Therefore, the disease similarity mining algorithm provided by the invention has certain accuracy and interpretation.
t-SNE algorithm: t-SNE (t-distributed stochastic neighbor embedding) is a machine learning algorithm for dimension reduction, proposed by Laurens van der Maaten and Geofrey Hinton in 08. the t-SNE is a nonlinear dimension reduction algorithm, and is very suitable for high-dimensional dimension reduction to 2-dimensional or 3-dimensional visualization. In practical applications, t-SNE is rarely used for dimension reduction, and is mainly used for visualization.
Preferably, the pharmaceutical knowledge graph comprises main components of medicines, treatment sites, treatment disease information, ATC coding and ICD10 coding information.
Preferably, the disease knowledge graph comprises treatment of disease, disease sites, common symptoms, ICD10 codes and drug and disease treatment relations.
Preferably, the main components, the treatment parts and the treatment disease information of the medicine can be obtained through the instruction book of the medicine; the ICD10 coding information can be supplemented by ICD10 coding for treating diseases; the ATC code can be obtained by reasoning the main components of the medicine; information such as treatment, disease incidence, common symptoms, ICD10 codes and the like of the diseases can be extracted through disease encyclopedia description; the therapeutic relationship of drugs to diseases is available through a number of electronic prescriptions.
The foregoing describes preferred embodiments of the present invention, but is not intended to limit the invention thereto. Modifications and variations to the embodiments disclosed herein may be made by those skilled in the art without departing from the scope and spirit of the invention.
Claims (1)
1. A method for mining similar diseases, which is characterized in that the method for mining similar diseases is applied to a prescription auditing system;
the method for excavating similar diseases comprises the following steps:
step one, constructing a knowledge graph: the knowledge graph comprises a pharmaceutical knowledge graph and a disease knowledge graph;
step two, obtaining vector representation of the disease based on the knowledge graph: in order to obtain vector representation of the disease, a plurality of node sequences are obtained in a random walk mode; taking the disease as an initial node, acquiring a next-hop node according to the connection relation among the nodes, and the like; generating a vector representation of the disease by a Word2Vec model after the plurality of node sequences are acquired; simultaneously adopting a t-SNE algorithm to reduce the vector representation of the high-dimensional diseases to a 2-dimensional vector, and drawing the vector representation on a 2-D coordinate system to represent the distance between the two diseases;
step three, evaluating similarity based on vector representation of the disease: calculating cosine similarity of two diseases according to vector representation of the diseases, wherein the numerical value of the cosine similarity is the similarity between the two diseases; wherein,
the pharmaceutical knowledge graph comprises main components of medicines, treatment parts, treatment disease information, ATC coding and ICD10 coding information; the disease knowledge graph comprises treatment of diseases, disease parts, common symptoms, ICD10 codes and treatment relations between medicines and diseases; the main components, the treatment parts and the treatment disease information of the medicine are obtained through the instruction book of the medicine; the ICD10 coding information is supplemented by ICD10 coding for treating diseases; the ATC code is obtained by reasoning main components of the medicine; treatment of disease, site of onset, common symptoms, ICD10 codes are extracted from disease encyclopedia; the treatment relationship between the medicine and the disease is obtained through a plurality of electronic prescriptions;
the vector representation of the disease generated by the Word2Vec model is divided into two steps of node sequence generation and disease vector calculation;
the generation of the node sequence specifically means that the node sequence is constructed by random walk on a knowledge graph; wherein, the random walk on the knowledge graph specifically refers to random walk according to a meta-path; the meta-path refers to a path which can be interpreted in medicine; the meta path specifically includes: 1) Drug-disease; 2) Drug-symptom-disease; 3) Medicine-principal component-disease; 4) Medicine-site-disease; 5) drug-ICD 10 code-disease; the nodes sampled by different element paths are spliced through the same node to generate a sequence, and different medicines or nodes of the diseases are spliced into a sequence through disease nodes or medicine nodes;
in the disease vector calculation, the input of the disease vector is a plurality of node sequences generated based on random walk of the knowledge graph; when the node sequence comprises a plurality of disease nodes which are related by drug nodes and symptom nodes, the related disease nodes have similar medical performances, namely can be treated by similar drugs or comprise the same similar symptoms; generating vector representation of each node in the knowledge graph by adopting a universal Word2Vec algorithm;
the method for mining similar diseases is applied to a multi-drug multi-symptom prescription traditional Chinese medicine symptom relation mining scene or an extended medicine symptom relation scene; in a multi-drug multi-symptom prescription traditional Chinese medicine relation mining scene, evaluating the most similar diagnosis of indications and prescriptions of medicine treatment based on the similar disease mining method; in the scenario of expanding the medicine relation, the method for mining similar diseases expands the medicine relation by clustering the diseases.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211679383.0A CN115662647B (en) | 2022-12-27 | 2022-12-27 | Method for excavating similar diseases and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211679383.0A CN115662647B (en) | 2022-12-27 | 2022-12-27 | Method for excavating similar diseases and application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115662647A CN115662647A (en) | 2023-01-31 |
| CN115662647B true CN115662647B (en) | 2024-03-12 |
Family
ID=85023271
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211679383.0A Active CN115662647B (en) | 2022-12-27 | 2022-12-27 | Method for excavating similar diseases and application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115662647B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109325131A (en) * | 2018-09-27 | 2019-02-12 | 大连理工大学 | A kind of drug identification method based on biomedical knowledge map reasoning |
| CN109616169A (en) * | 2018-12-24 | 2019-04-12 | 广州天鹏计算机科技有限公司 | Similar patients' method for digging, device, computer equipment and storage medium |
| CN110675954A (en) * | 2019-10-11 | 2020-01-10 | 北京百度网讯科技有限公司 | Information processing method and device, electronic equipment and storage medium |
| CN112151184A (en) * | 2020-09-27 | 2020-12-29 | 东北林业大学 | System for calculating disease similarity based on network representation learning |
| CN112309565A (en) * | 2020-08-28 | 2021-02-02 | 北京京东世纪贸易有限公司 | Method, apparatus, electronic device, and medium for matching drug information and disorder information |
| CN112542243A (en) * | 2020-12-05 | 2021-03-23 | 大连东软教育科技集团有限公司 | ICU electronic medical record knowledge graph construction method, system and storage medium |
| CN113539409A (en) * | 2021-07-28 | 2021-10-22 | 平安科技（深圳）有限公司 | Treatment scheme recommendation method, device, equipment and storage medium |
| CN113628726A (en) * | 2021-08-10 | 2021-11-09 | 海南榕树家信息科技有限公司 | Traditional Chinese medicine dialectical treatment recommendation system and method based on graph neural network and electronic equipment |
| CN114510579A (en) * | 2022-02-08 | 2022-05-17 | 吾征智能技术(北京)有限公司 | Disease automatic question-answering system, equipment and medium based on medical knowledge map |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101721528B1 (en) * | 2015-05-28 | 2017-03-31 | 아주대학교산학협력단 | Method for providing disease co-occurrence probability from disease network |
-
2022
- 2022-12-27 CN CN202211679383.0A patent/CN115662647B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109325131A (en) * | 2018-09-27 | 2019-02-12 | 大连理工大学 | A kind of drug identification method based on biomedical knowledge map reasoning |
| CN109616169A (en) * | 2018-12-24 | 2019-04-12 | 广州天鹏计算机科技有限公司 | Similar patients' method for digging, device, computer equipment and storage medium |
| CN110675954A (en) * | 2019-10-11 | 2020-01-10 | 北京百度网讯科技有限公司 | Information processing method and device, electronic equipment and storage medium |
| CN112309565A (en) * | 2020-08-28 | 2021-02-02 | 北京京东世纪贸易有限公司 | Method, apparatus, electronic device, and medium for matching drug information and disorder information |
| CN112151184A (en) * | 2020-09-27 | 2020-12-29 | 东北林业大学 | System for calculating disease similarity based on network representation learning |
| CN112542243A (en) * | 2020-12-05 | 2021-03-23 | 大连东软教育科技集团有限公司 | ICU electronic medical record knowledge graph construction method, system and storage medium |
| CN113539409A (en) * | 2021-07-28 | 2021-10-22 | 平安科技（深圳）有限公司 | Treatment scheme recommendation method, device, equipment and storage medium |
| CN113628726A (en) * | 2021-08-10 | 2021-11-09 | 海南榕树家信息科技有限公司 | Traditional Chinese medicine dialectical treatment recommendation system and method based on graph neural network and electronic equipment |
| CN114510579A (en) * | 2022-02-08 | 2022-05-17 | 吾征智能技术(北京)有限公司 | Disease automatic question-answering system, equipment and medium based on medical knowledge map |
Non-Patent Citations (8)
| Title |
|---|
| 代涛 编.《医学信息学进展【2015—2020】》.中华医学电子音像出版社,2020,正文第118-121页. * |
| 余黄樱子 等.基于网络表示学习的疾病知识关联挖掘与预测方法研究.情报理论与实践.2019,第42卷(第12期),全文. * |
| 刘宏志 编.《推荐系统》.北京：机械工业出版,2020,正文第150-155页. * |
| 唐子惠 编.《医学人工智能导论》.上海：上海科学技术出版社,2020,正文第128-132页. * |
| 拉姆什·沙尔达 等著.《商业分析 基于数据科学及人工智能技术的决策支持系统 原书第11版》.北京：机械工业出版社,2022,正文第524-526页. * |
| 李杰 等.基于疾病本体的疾病相似性计算方法.生物化学与生物物理进展.2015,(第02期),全文. * |
| 王亦凡 等.基于异构图嵌入学习的相似病案推荐.计算机系统应用.2020,第29卷(第10期),全文. * |
| 王青青.融合知识图谱与图嵌入的电子病历推荐方法研究.《中国优秀硕士学位论文全文数据库 医药卫生科技辑(月刊)》.2021,(第12期),正文第1-40页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115662647A (en) | 2023-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109599185A (en) | Disease data processing method, device, electronic equipment and computer-readable medium | |
| Guala | Experimental localism and external validity | |
| Talati | AI in healthcare domain | |
| CN108717441B (en) | Method and device for determining predicates corresponding to problem templates | |
| Ali | Abdullah | |
| Robles-Bykbaev et al. | SPELTA: An expert system to generate therapy plans for speech and language disorders | |
| CN115130656A (en) | Training method, device and equipment of anomaly detection model and storage medium | |
| CN109636421A (en) | Medical data exception recognition methods, equipment and storage medium based on machine learning | |
| Sirigineedi et al. | The Early Detection Of Alzheimer's Illness Using Machine Learning And Deep Learning Algorithms | |
| Zehra et al. | Physiological and chaos effect on dynamics of neurological disorder with memory effect of fractional operator: A mathematical study | |
| CN115662647B (en) | Method for excavating similar diseases and application | |
| Pratl | Processing and symbolization of ambient sensor data | |
| CN109616165A (en) | Medical information methods of exhibiting and device | |
| Cornfield | A statistician's apology | |
| Liotine et al. | Managing a Smart City's Resilience to Ebola: An Ontological Framework | |
| CN113657086A (en) | Word processing method, device, equipment and storage medium | |
| Scott | Roberto Bolaño’s 2666: serial murder and narrative necrosis | |
| Ansari et al. | Adaptive neurofuzzy system for tuberculosis | |
| Gantenbein | Watson, come here! The role of intelligent systems in health care | |
| CN116403731A (en) | Missense mutation effect prediction method and system for clinical drug effect based on deep learning | |
| Rao et al. | AI-enabled clinical decision support system | |
| Ahmed et al. | Rock-paper-scissors image classification using transfer learning | |
| Jay et al. | Using formal concept analysis for mining and interpreting patient flows within a healthcare network | |
| Avdić et al. | Ensuring the Durability and Reliability of Data in Smart Health Services Using Blockchain Technologies | |
| CN112086206A (en) | Prescription searching method and device for deduction by using editable knowledge graph |
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 |