An AI-based elderly care system and method

CN121938642BActive Publication Date: 2026-05-29FUJIAN ZHONGWEIAN OCCUPATIONAL HEALTH ENG RES INST
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
FUJIAN ZHONGWEIAN OCCUPATIONAL HEALTH ENG RES INST
Filing Date
2026-03-27
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing technologies are difficult to adapt to the heterogeneous evolution of individual physiological characteristics in the monitoring of physiological indicators in the elderly, resulting in high-frequency invalid alarms or ignoring hidden organ function decline. Furthermore, traditional algorithms are difficult to meet the requirements of long-term monitoring in terms of computational overhead and stability when faced with fragmented physiological signal morphology or random nursing interventions.

Method used

By constructing dynamic topological associations of multimodal physiological time series, combining adaptive baseline evolution mechanisms and nursing intervention label purification logic, and using local sensitive hash function clusters to map to fixed-length hash retrieval codes, Hamming distance is calculated and structured nursing decision data packages are generated.

Benefits of technology

It achieves high-fidelity assessment of the health status of the elderly, reduces false positive alarms, and improves the stability of the system under extremely complex conditions and the feasibility of predictive nursing interventions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121938642B_ABST
    Figure CN121938642B_ABST
Patent Text Reader

Abstract

The application relates to the technical field of medical care information science, and discloses an AI-based old-age care system and method, which comprises the following steps: a signal acquisition module acquires physiological signals of an audience; a feature segmentation module extracts physiological feature vectors by using a physiological feature vector library; a topological modeling module constructs a physiological state directed topological structure; a search matching module maps the physiological topological structure into a binary hash search code by using a local sensitive hash function cluster, and matches a reference topological feature in a physiological normal state library to determine a Hamming distance; and a care instruction generation module determines a physiological abnormal state according to the Hamming distance and generates a care decision data packet. In the application, waveform evolution is converted into a topological atlas, and a high-dimensional search mechanism is introduced, so that the false alarm problem caused by physiological parameter dynamic drift is effectively overcome, and the reliability of the identification of hidden recession is improved.
Need to check novelty before this filing date? Find Prior Art