A medical equipment intelligent early warning method and system based on multi-modal data fusion

CN122337537APending Publication Date: 2026-07-03SHANXI CANCER HOSPITAL
View PDF -1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANXI CANCER HOSPITAL
Filing Date
2026-03-05
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

The alarm functions of existing medical monitoring equipment are susceptible to external interference, leading to false alarms. They lack predictability, and cloud computing causes response delays. Furthermore, they fail to combine individual patient physiological characteristics with clinical pathological rules, resulting in insufficient accuracy in early warning.

Method used

By employing an embedded AI chip and using a multimodal data fusion method, physiological data is acquired and preprocessed to generate personalized baselines. A fusion inference model is used to quantify the risk of pathological correlation, and combined with a clinical rule base to output graded early warnings, thus achieving early composite early warning.

Benefits of technology

It reduces false alarms, improves the accuracy and response speed of early warnings, is suitable for intensive care scenarios, and realizes the upgrade from data acquisition terminal to intelligent early warning terminal.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122337537A_ABST
    Figure CN122337537A_ABST
Patent Text Reader

Abstract

This invention provides an intelligent early warning method and system for medical devices based on multimodal data fusion. The method includes: acquiring multimodal physiological data from medical monitoring devices; the multimodal physiological data being timestamped; preprocessing the multimodal physiological data; generating an initial baseline from the preprocessed data using a clustering algorithm, and then adjusting the initial baseline by combining it with a pre-built clinical parameter correction library to obtain a personalized baseline; using a pre-trained fusion inference model, combining the input multimodal features with baseline deviation data for inference, and outputting a pathological correlation quantified risk score; comparing the pathological correlation quantified risk score with a preset threshold, and outputting a graded early warning result based on the comparison result. Its beneficial effects are: utilizing the fusion inference model to process the collected multimodal physiological data to achieve early composite early warning, and also reducing the defects of delay and insufficient early warning accuracy.
Need to check novelty before this filing date? Find Prior Art