CN-122025053-A - Knowledge graph-based quality evaluation method and system for medical records of traditional Chinese medicine

CN122025053ACN 122025053 ACN122025053 ACN 122025053ACN-122025053-A

Abstract

The invention discloses a traditional Chinese medicine medical record quality evaluation method and system based on a knowledge graph, which particularly relate to the field of medical record quality evaluation, and are characterized in that explicit association paths between syndrome diagnosis and prescription medication are extracted through analyzing traditional Chinese medicine medical record text data to form explicit association data, constraint matching analysis is carried out on logic consistency between the syndrome diagnosis, treatment method selection and prescription medication composition to obtain an explicit consistency evaluation result, semantic recognition is carried out on medication rationality interpretation information to obtain an abnormal medication rationality interpretation evaluation result, finally the explicit consistency evaluation result and the abnormal medication rationality interpretation evaluation result are fused to distinguish abnormal medication situations to form a medical record logic rationality evaluation result, quality judgment is carried out on traditional Chinese medicine medical histories according to the medical record logic rationality evaluation result, and the traditional Chinese medicine medical record quality evaluation result is output.

Inventors

CUI SHUKE
CUI HAIYANG
LIU CHENGCANG
WEI LINYOU
MIAO JING
CUI HAN

Assignees

河南经方云科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. The quality evaluation method of the medical record of the traditional Chinese medicine based on the knowledge graph is characterized by comprising the following steps: S1, acquiring text data of a traditional Chinese medical record, and analyzing medical record text elements in the text data of the traditional Chinese medical record to construct a basic data set of the traditional Chinese medical record; S2, constructing a syndrome-treatment-prescription drug association knowledge graph based on a traditional Chinese medical record basic data set, and extracting a dominant association path between syndrome diagnosis and prescription drug to form dominant association data; S3, based on the dominant correlation data, carrying out constraint matching analysis on the logical consistency between the diagnosis and treatment method selection and the prescription composition, and generating a dominant consistency assessment result; s4, carrying out semantic recognition on medication rationality interpretation information based on a basic data set of the medical record of the traditional Chinese medicine, analyzing rationality basis of prescription medication deviating from a dominant association path, and generating an abnormal medication rationality interpretation evaluation result; s5, carrying out fusion analysis on the dominant consistency evaluation result and the abnormal medication rationality interpretation evaluation result, and distinguishing abnormal medication situations to form a medical record logic rationality evaluation result; And S6, carrying out quality judgment on the traditional Chinese medical calendar according to the medical record logic rationality evaluation result, and outputting a traditional Chinese medical record quality evaluation result.
2. The method for evaluating the quality of a traditional Chinese medical record based on a knowledge graph according to claim 1, wherein the medical record text elements comprise disease diagnosis, syndrome diagnosis, treatment description, prescription composition, medication instruction and explanatory sentences.
3. The knowledge-graph-based quality evaluation method for the medical record of traditional Chinese medicine according to claim 2, wherein the method is characterized in that the method S1 specifically comprises the following steps: Obtaining text data of a medical record of a traditional Chinese medicine, and performing character standardization, clause processing and field positioning on the text data of the medical record of the traditional Chinese medicine to form a disease diagnosis text segment, a evidence waiting diagnosis text segment, a treatment description text segment, a prescription composition text segment, a medication description text segment and an explanatory sentence text segment; Performing term normalization and entity extraction on each text segment to generate a disease diagnosis item, a certificate waiting diagnosis item, a treatment item, a prescription item and an explanation item; And constructing syndrome nodes, treatment nodes, prescription nodes and interpretation nodes based on the item correspondence, and establishing node indexes to obtain a basic data set of the medical record of the traditional Chinese medicine.
4. The knowledge-graph-based quality evaluation method for traditional Chinese medical records of claim 3, wherein S2 specifically comprises: Based on a traditional Chinese medical record basic data set, extracting the associated edges of the syndrome node and the rule node and the associated edges of the rule node and the prescription node according to the node indexes of the syndrome node, the rule node, the prescription node and the interpretation node, and writing the extracted edges into a graph structure to form a syndrome-rule-prescription association knowledge graph; And in the syndrome-therapy-prescription drug correlation knowledge graph, path traversal is performed along the reachable path from the syndrome node to the prescription drug node, the dominant correlation path between syndrome diagnosis and prescription drug is extracted, and syndrome-therapy-prescription drug dominant correlation data is output.
5. The knowledge-graph-based quality evaluation method for the medical record of traditional Chinese medicine according to claim 4, wherein the step S3 is specifically as follows: based on the dominant association data of syndrome-treatment-prescription, converting the dominant association path into a syndrome diagnosis-treatment selection-prescription composition association record according to the node index corresponding relation among syndrome nodes, treatment nodes and prescription nodes; Establishing rule constraint items according to preset Chinese medicine dialectical treatment standard rules, executing condition verification, and outputting matching marks, unmatched reason items and rule triggering items of the associated records; Summarizing the matching mark and the rule triggering item to generate a syndrome-treatment-prescription dominant consistency evaluation result.
6. The method for evaluating the quality of a medical record of traditional Chinese medicine based on a knowledge graph according to claim 5, wherein the medication rationality interpretation information comprises constitution description, concurrent evidence description, concurrent diagnosis description and medication contraindication.
7. The knowledge-graph-based quality evaluation method for the medical record of traditional Chinese medicine according to claim 6, wherein the step S4 is specifically as follows: based on the basic data set of the medical record of the traditional Chinese medicine, locating an explanatory sentence text segment according to node indexes of explanatory nodes and extracting explanatory items; Executing semantic recognition based on the explanation items to form constitution description, double-duty description, combined diagnosis description and medication tabu description; Positioning prescription item corresponding to prescription based on syndrome-treatment-prescription dominant correlation data and calculating deviation information deviating from the dominant correlation path; And generating an abnormal medication rationality interpretation evaluation result according to the corresponding relation between the deviation information and the constitution description, the concurrent inclusion syndrome description, the combined diagnosis description and the medication tabu description.
8. The knowledge-graph-based quality evaluation method for the medical record of traditional Chinese medicine according to claim 7, wherein the step S5 is specifically as follows: Establishing a mapping relation between an unmatched reason item and a rule triggering item in a syndrome-treatment-prescription dominant consistency evaluation result and deviation information and constitution description, double-evidence description, combined diagnosis description and medication tabu description in an abnormal medication rationality interpretation evaluation result; Generating consistency scores and interpretation coverage marks according to the prescription item alignment mapping relation; and outputting abnormal medication situation classification labels according to the consistency scores and the interpretation coverage marks to form a medical record logical rationality assessment result.
9. The knowledge-graph-based quality evaluation method for the medical record of the traditional Chinese medicine, according to claim 8, is characterized in that S6 specifically comprises: performing quality classification judgment on the disease diagnosis items, the evidence waiting diagnosis items, the treatment items and the prescription items in the traditional Chinese medicine calendar basic data set according to the consistency scores and the interpretation coverage marks in the medical record logical rationality evaluation results; And based on the quality classification judgment result, determining high-quality medical records, reasonable exceptional medical records and non-standard medical record class labels, and outputting a quality evaluation result of the traditional Chinese medical records.
10. A knowledge-graph-based quality evaluation system for a medical record of traditional Chinese medicine, for implementing the knowledge-graph-based quality evaluation method for the medical record of traditional Chinese medicine according to any one of claims 1 to 9, comprising: the medical record analysis module is used for acquiring text data of the traditional Chinese medical record, analyzing medical record text elements in the text data of the traditional Chinese medical calendar and constructing a basic data set of the traditional Chinese medical record; the pattern construction module is used for constructing a syndrome-treatment-prescription drug association knowledge pattern based on a traditional Chinese medical record basic data set, extracting a dominant association path between syndrome diagnosis and prescription drug, and forming dominant association data; the consistency verification module is used for carrying out constraint matching analysis on the logical consistency between the diagnosis and treatment method selection and the prescription composition based on the dominant correlation data to generate a dominant consistency assessment result; The interpretation and identification module is used for carrying out semantic identification on medication rationality interpretation information based on a basic data set of the medical record of the traditional Chinese medicine, analyzing rationality basis of deviation of prescription medication from a dominant association path and generating an abnormal medication rationality interpretation and evaluation result; The fusion judging module is used for carrying out fusion analysis on the dominant consistency evaluation result and the abnormal medication rationality interpretation evaluation result, and distinguishing abnormal medication situations to form a medical record logic rationality evaluation result; and the quality judgment module is used for judging the quality of the traditional Chinese medicine calendar according to the medical record logic rationality evaluation result and outputting the quality evaluation result of the traditional Chinese medicine medical record.

Description

Knowledge graph-based quality evaluation method and system for medical records of traditional Chinese medicine Technical Field The invention relates to the technical field of medical record quality evaluation, in particular to a method and a system for evaluating quality of traditional Chinese medical record based on a knowledge graph. Background The quality evaluation work of the current clinical medical records of the traditional Chinese medicine mainly depends on manual examination and combines the personal experience of doctors to judge the normalization and rationality of the medical records. The method has the problems of strong subjectivity, difficult unification of auditing standards, low evaluation efficiency and the like. In actual work, the medical record quality evaluation standards among different auditors are difficult to agree, so that the reliability and stability of the medical record evaluation result are insufficient, and the overall quality of the clinical medical record is difficult to be effectively ensured. The existing traditional Chinese medicine medical record quality evaluation method lacks objective and unified evaluation basis, so that the accuracy and consistency of medical record quality evaluation results are insufficient. In order to solve the above problems, a technical solution is now provided. Disclosure of Invention In order to overcome the defects in the prior art, the embodiment of the invention provides a method and a system for evaluating quality of a medical record of traditional Chinese medicine based on a knowledge graph, which are used for solving the problems in the background art. In order to achieve the above purpose, the present invention provides the following technical solutions: a quality evaluation method of a traditional Chinese medical record based on a knowledge graph comprises the following steps: S1, acquiring text data of a traditional Chinese medical record, and analyzing medical record text elements in the text data of the traditional Chinese medical record to construct a basic data set of the traditional Chinese medical record; S2, constructing a syndrome-treatment-prescription drug association knowledge graph based on a traditional Chinese medical record basic data set, and extracting a dominant association path between syndrome diagnosis and prescription drug to form dominant association data; S3, based on the dominant correlation data, carrying out constraint matching analysis on the logical consistency between the diagnosis and treatment method selection and the prescription composition, and generating a dominant consistency assessment result; s4, carrying out semantic recognition on medication rationality interpretation information based on a basic data set of the medical record of the traditional Chinese medicine, analyzing rationality basis of prescription medication deviating from a dominant association path, and generating an abnormal medication rationality interpretation evaluation result; s5, carrying out fusion analysis on the dominant consistency evaluation result and the abnormal medication rationality interpretation evaluation result, and distinguishing abnormal medication situations to form a medical record logic rationality evaluation result; And S6, carrying out quality judgment on the traditional Chinese medical calendar according to the medical record logic rationality evaluation result, and outputting a traditional Chinese medical record quality evaluation result. In a preferred embodiment, the medical record text elements include disease diagnosis, syndrome diagnosis, therapy description, prescription composition, instructions for administration, and explanatory statements. In a preferred embodiment, S1 is specifically: Obtaining text data of a medical record of a traditional Chinese medicine, and performing character standardization, clause processing and field positioning on the text data of the medical record of the traditional Chinese medicine to form a disease diagnosis text segment, a evidence waiting diagnosis text segment, a treatment description text segment, a prescription composition text segment, a medication description text segment and an explanatory sentence text segment; Performing term normalization and entity extraction on each text segment to generate a disease diagnosis item, a certificate waiting diagnosis item, a treatment item, a prescription item and an explanation item; And constructing syndrome nodes, treatment nodes, prescription nodes and interpretation nodes based on the item correspondence, and establishing node indexes to obtain a basic data set of the medical record of the traditional Chinese medicine. In a preferred embodiment, S2 is specifically: Based on a traditional Chinese medical record basic data set, extracting the associated edges of the syndrome node and the rule node and the associated edges of the rule node and the prescription node according to the node indexes of the syndrome no