Wearable device-based out-of-hospital monitoring and warning system for aura-laden seizures

CN122123652APending Publication Date: 2026-06-02THE THIRD PEOPLES HOSPITAL OF CHENGDU

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Applications(China)
Current Assignee / Owner: THE THIRD PEOPLES HOSPITAL OF CHENGDU
Filing Date: 2026-03-05
Publication Date: 2026-06-02

AI Technical Summary

Technical Problem

Existing wearable devices lack effective identification of brain electrical activity during the prodromal phase in epileptic seizure monitoring, resulting in high false alarm and false negative rates. They also lack personalized dynamic threshold settings and multimodal data fusion capabilities, have incomplete emergency call procedures, and their positioning function is not integrated with the early warning process, making it difficult to meet the needs for reliable monitoring in all scenarios and around the clock.

Method used

A unified clock protocol is used to synchronize the signal marking device, heart rate monitoring module, and positioning module. Static calibration and personalized threshold settings are performed. The risk of attack is assessed in real time based on a dynamic probability recursive model. A multi-level emergency call process and log management are designed.

Benefits of technology

It achieves synchronous acquisition of multimodal data with millisecond-level precision, eliminates noise and outliers, adaptively sets thresholds, reduces false alarm rate, improves early warning accuracy and rescue efficiency, and ensures the reliability and traceability of emergency calls.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN122123652A_ABST

Patent Text Reader

Abstract

This invention relates to the field of outpatient monitoring and early warning technology for epileptic seizures with aura, and discloses an outpatient monitoring and early warning system for epileptic seizures with aura based on wearable devices. First, a VEEG marker, heart rate monitor, and positioning module are fixed to the wrist. Multiple device clocks are activated and synchronized, and buffers are allocated. A calibration interval is set, and the mean and standard deviation of heart rate are statistically analyzed, with samples exceeding the limit being removed. A 300-second baseline is collected, and the number of auras and average heart rate are statistically analyzed in segments at equal time intervals. Aura and heart rate thresholds are adaptively extracted using the upper quartiles. An initial prior is set, and the feature ratio and normalized likelihood of each segment are calculated. The posterior probability is recursively calculated, and a judgment threshold is determined. Features are extracted from real-time data segments, and the posterior probability is recursively calculated online. An early warning is triggered when the threshold is exceeded. After an early warning, a multi-level call to primary and secondary contacts is automatically initiated, and detailed logs are recorded.

Need to check novelty before this filing date? Find Prior Art