Intelligent processing method and system for clinical research data

By analyzing the correlation and influence between data subsets and reference subsets in a multidimensional index dataset, and adaptively adjusting clustering parameters, the problem of low accuracy in identifying outlier data caused by improper selection of clustering parameters in existing technologies is solved, achieving higher identification accuracy and analysis reliability.

CN121641492BActive Publication Date: 2026-06-09SHANGHAI JIAOTONG UNIV SCHOOL OF MEDICINE
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511826614.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-12-05
Publication Date
2026-06-09
Estimated Expiration
2045-12-05

AI Technical Summary

Technical Problem

Existing methods for identifying abnormal data in clinical research data often suffer from low accuracy due to improper selection of clustering parameters, making it difficult to effectively distinguish between individual patient differences and genuinely erroneous data.

Method used

By obtaining the degree of correlation and influence between each data subset and the reference subset in the multidimensional indicator dataset, the clustering parameters are adaptively adjusted, and methods such as principal component analysis, scatter plot fitting, and DBSCAN clustering are used to identify outlier data.

Benefits of technology

It improved the accuracy of abnormal data identification, reduced the impact of individual patient differences on the results, and enhanced the reliability of the analysis results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121641492B_ABST
    Figure CN121641492B_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of data processing, and particularly relates to a kind of wisdom processing method and system of clinical research data, obtain the multidimensional index data group of each patient and form multidimensional index data set, and it is divided into at least two data subsets according to dimension, for any data subset, obtain the reference subset of any data subset, according to the data distribution characteristics in each reference subset, and the data variation characteristics between any data subset and each reference subset, obtain the correlation degree between any data subset and each reference subset;According to the correlation degree between any data subset and each reference subset, obtain the adaptive clustering parameter of each data point when clustering to any data subset;Using the adaptive clustering parameter of each data point when clustering to each data subset, each data subset is clustered, and the abnormal index data in each data subset is obtained, to improve the accuracy of abnormal data identification.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Online identification method for atmospheric pollution

    CN115659195A

  • Wind power plant data quality evaluation method based on high-dimensional space anomaly identification

    CN119249324A