CN-121981604-A - Intelligent evaluation method and system based on practice capability of clinical study coordinator

CN121981604ACN 121981604 ACN121981604 ACN 121981604ACN-121981604-A

Abstract

The invention discloses an intelligent evaluation method and system based on the practical capability of a clinical research coordinator, which relates to the technical field of artificial intelligence application, the method comprises the steps of collecting multi-source data related to the practical capability of a clinical study coordinator, performing standardized processing on the multi-source data to eliminate data format differences and content redundancy, obtaining a structured data set and the like. According to the invention, the multi-dimensional semantic association of evaluation data, capability indexes, test stages, post responsibilities and case scenes is realized by constructing the case knowledge graph, a quantitative evaluation system is simultaneously established, and the dynamic evaluation mechanism and the personalized training path generation function are combined, so that the problems that the judgment is dependent on manual experience, a unified objective quantitative system is lacking, the real practice capability is difficult to reflect by a fixed mode, the evaluation data is scattered and uncorrelated, a structured knowledge system cannot be formed and personalized training analysis cannot be formed are solved, and the CRC capability evaluation is updated from experience driving to data driving and from static unification to dynamic individuation.

Inventors

XU JUN
Huang sichao
HUANG KEER
HU HUAZHONG
LIU HANGBO
SUN PINGHUA
ZHOU HAIBO
Tan Qiutong
Mo Enpan
CHEN LIN
CHEN WENYING
FANG RUI

Assignees

暨南大学
广州科泰医药科技有限公司
武汉幻之路信息科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. An intelligent assessment method based on clinical study coordinator practical ability is characterized by comprising the following steps: s1, collecting multi-source data related to practical capability of a clinical study coordinator, and performing standardized processing on the multi-source data to eliminate data format differences and content redundancy, so as to obtain a structured data set; S2, constructing a case knowledge graph based on the structured data set to form a multi-dimensional semantic knowledge network of capability elements such as coverage capability indexes, post responsibilities, test phases and the like and clinical practice scenes; S3, dynamically evaluating the practical capability of a clinical research coordinator in different test stages based on the case knowledge graph to generate a quantized capability evaluation result; s4, identifying weak ability items of clinical study coordinators according to the quantitative ability evaluation result, matching case resources corresponding to the weak ability items with standard operation contents, and generating personalized training paths; And S5, acquiring training feedback data generated by a clinical research coordinator based on the personalized training path and newly-added clinical practice case information, and updating node attributes and semantic weights of the case knowledge graph to form an evaluation-training-feedback-re-evaluation self-learning closed loop.
2. The intelligent assessment method based on clinical study coordinator practice ability according to claim 1, wherein the step of collecting multi-source data related to clinical study coordinator practice ability in S1, performing normalization processing on the multi-source data to eliminate data format differences and content redundancy, comprises: collecting clinical study coordinator practice core ability evaluation index system data and standardized clinical practice case data; The standardized clinical practice case data comprise case numbers, test stages, post responsibilities, problem descriptions, solutions, correction results and jeopardy contents; The standardized processing comprises repeated item detection, missing field completion, semantic field mapping and outlier detection.
3. The intelligent assessment method based on clinical study coordinator practical ability according to claim 1, wherein in S2, a case knowledge graph is constructed based on the structured data set, and a multidimensional semantic knowledge network covering ability factors such as ability index, post responsibility, test phase and the like and clinical practice scene is formed, which comprises: Defining entity types and semantic association by adopting a node-relation model, wherein the entity types comprise test items, capability indexes, test stages, post responsibilities, cases, problems, solutions, correction results and dislikes; the semantic association comprises a source relation between a case and a test item, a corresponding relation between the case and a capability index and an adopted relation between a problem and a solution; And automatically generating nodes and relations through a Neo4j graph database and a Py2Neo interface, and optimizing the node space distribution by calling a force-directed layout algorithm to form the multidimensional semantic knowledge network.
4. The intelligent assessment method based on clinical study coordinator practical ability according to claim 1, wherein in S3, based on the case knowledge graph, the intelligent assessment method dynamically assesses the practical ability of the clinical study coordinator in different test phases, comprising: Searching a correlation path corresponding to the capability index in the case knowledge graph through a Cypher query statement, synchronously correlating entity nodes such as a test stage, a post responsibility and the like to which the capability index belongs, and extracting a complete behavior chain of a problem-solving method-rectifying result; Taking node association degree, path integrity and semantic consistency as evaluation parameters, realizing quantitative calculation through Python, and generating the quantitative capability evaluation result; wherein a STREAMLIT framework and ECharts component are employed to enable dynamic visual presentation of the quantitative capability assessment results.
5. The intelligent assessment method based on the practice ability of the clinical study coordinator according to claim 4, wherein the process of dynamically assessing the practice ability of the clinical study coordinator in different trial phases further comprises generating assessment questions, in particular: Extracting a case associated with the capability index from the case knowledge graph, converting the problem description in the case into an evaluation problem stem, and designing correct answers and error options of the evaluation problem according to the solution method and the correction result in the case.
6. The intelligent assessment method based on clinical study coordinator practical ability according to claim 1, wherein identifying the ability Bao Ruoxiang of the clinical study coordinator and generating a personalized training path according to the quantitative ability assessment result in S4 comprises: Based on the semantic similarity between the capability Bao Ruoxiang and the case knowledge graph, matching the case resource corresponding to the weak capability item with the standard operation content; The personalized training path comprises a question-by-question comment which comprises correct answers, error reasons, standard operation guide and boundary condition explanation; and generating a visual achievement trend curve and a learning priority prompt through the quantitative capability assessment result association analysis.
7. The intelligent assessment method based on clinical study coordinator practical ability according to claim 6, wherein matching the case resource corresponding to the weak ability item with the normative operation content based on the semantic similarity between the ability Bao Ruoxiang and the case knowledge graph comprises: Aiming at the identified weak items of the ability, searching case nodes which are directly related to the ability indexes corresponding to the weak items of the ability in the case knowledge graph through a Cypher query statement, and screening cases with consistent test stage and post responsibilities which are conventionally related to the ability indexes; And extracting a solution, a correction result and a dislike content from the retrieved case nodes, integrating the solution, the correction result and the dislike content into training materials, and forming training content corresponding to the weak ability items.
8. The intelligent assessment method based on clinical study coordinator practical ability according to claim 1, wherein updating node attributes and semantic weights of the case knowledge graph in S5 comprises: training feedback data and newly added case information are processed through a semantic aggregation algorithm, node attributes are automatically supplemented, and semantic weights are adjusted; Performing logic consistency verification on the newly added semantic association, and writing the case knowledge graph after the verification is passed; And simultaneously recording a high-frequency problem and a retrieval path, and optimizing semantic index precision and reasoning logic of the case knowledge graph.
9. The intelligent assessment method based on clinical study coordinator practical ability according to claim 1, wherein the method further comprises a cross-project adaptation step comprising: identifying semantic correspondence among different test projects through a unified clinical study coordinator practice ability index standard system and a semantic layer data model; the fusion analysis and training content multiplexing of the data related to the practical ability of the multi-center and multi-project clinical research coordinator are realized, and the rapid deployment of different types of medicines and medical instrument clinical test scenes is supported.
10. An intelligent assessment system based on clinical study coordinator practice ability, comprising: The data acquisition and management module is used for acquiring multi-source data related to the practical capability of a clinical research coordinator, and performing standardized processing on the multi-source data to eliminate data format differences and content redundancy and obtain a structured data set; The case knowledge graph construction module is used for constructing a case knowledge graph based on the structured data set to form a multi-dimensional semantic knowledge network of capability factors such as coverage capability indexes, post responsibilities, test phases and the like and clinical practice scenes; The intelligent evaluation analysis module is used for dynamically evaluating the practical ability of a clinical research coordinator in different test stages based on the case knowledge graph to generate a quantized ability evaluation result; The personalized training recommendation module is used for identifying weak items of the ability of clinical study coordinators according to the quantized ability evaluation result, matching case resources corresponding to the weak items with standard operation contents and generating a personalized training path; And the self-learning and feedback optimization module is used for collecting training feedback data generated by a clinical research coordinator based on the personalized training path and newly-added clinical practice case information, updating node attributes and semantic weights of the case knowledge graph, and forming a self-learning closed loop of evaluation, training, feedback and reevaluation.

Description

Intelligent evaluation method and system based on practice capability of clinical study coordinator Technical Field The invention relates to the technical field of artificial intelligence application, in particular to an intelligent evaluation method and system based on the practical capability of a clinical research coordinator. Background Clinical study coordinator (CRC) is taken as an execution post in the whole clinical test process of medicines and medical instruments, bears key responsibilities throughout the test, and specifically covers multiple works such as subject recruitment screening, group entry management, visit process organization and execution, test data acquisition, check and report, test material management, ethical file archiving and updating, auxiliary report and tracking of adverse events and scheme deviation, and communication coordination among researchers, subjects, ethical committees, sponsors and the like. The practical core capability of CRC directly relates to the authenticity, integrity and accuracy of clinical test data, influences the compliance execution of a test scheme, is a key factor for guaranteeing the rights and the safety of a subject, reducing the test risk and promoting the smooth progress of a research project, and has become one of important indexes for measuring the quality of clinical tests. At present, CRC capability assessment and training mainly depend on manual experience judgment or an offline teaching mode, a unified and objective quantification system is lacking, the existing mode is mainly based on fixed question banks, paper records or static scoring, true practice capability of the CRC in different test stages and different task scenes is difficult to reflect, meanwhile, assessment data is stored in a scattered mode and lacks semantic association, a structured knowledge system cannot be formed, and data-based individualized training analysis cannot be realized. Disclosure of Invention In order to overcome the above-mentioned drawbacks of the prior art, the present invention provides an intelligent assessment method and system based on the practical capabilities of a clinical research coordinator to solve the problems set forth in the background art. In a first aspect, an embodiment of the present application provides an intelligent evaluation method based on a clinical study coordinator's practical ability, which is characterized by comprising: s1, collecting multi-source data related to practical capability of a clinical study coordinator, and performing standardized processing on the multi-source data to eliminate data format differences and content redundancy, so as to obtain a structured data set; S2, constructing a case knowledge graph based on the structured data set to form a multi-dimensional semantic knowledge network of capability elements such as coverage capability indexes, post responsibilities, test phases and the like and clinical practice scenes; S3, dynamically evaluating the practical capability of a clinical research coordinator in different test stages based on the case knowledge graph to generate a quantized capability evaluation result; s4, identifying weak items of the ability of a clinical study coordinator according to the quantitative ability assessment result, matching case resources and standard operation contents corresponding to the weak items, and generating personalized training paths; And S5, acquiring training feedback data generated by a clinical research coordinator based on the personalized training path and newly added clinical practice case information, and updating node attributes and semantic weights of a case knowledge graph to form an evaluation-training-feedback-reevaluation self-learning closed loop. Optionally, collecting multi-source data related to clinical study coordinator practice ability in S1, performing normalization processing on the multi-source data to eliminate data format differences and content redundancy, including: collecting clinical study coordinator practice core ability evaluation index system data and standardized clinical practice case data; the standardized clinical practice case data comprise case numbers, test stages, post responsibilities, problem descriptions, solutions, correction results and jeopardy contents; the normalization processing comprises duplicate term detection, missing field completion, semantic field mapping and outlier detection. Optionally, in S2, constructing a case knowledge graph based on the structured data set, to form a multidimensional semantic knowledge network for covering capability factors such as capability indexes, post responsibilities, test phases, and the like, and clinical practice scenes, including: Defining entity types and semantic association by adopting a node-relation model, wherein the entity types comprise test items, capability indexes, test stages, post responsibilities, cases, problems, solutions, correction results and returnables;