Cardiac sudden death risk early warning method based on coupling analysis of electrocardio and sleep signals
By synchronously acquiring ECG and sleep signals, and utilizing a parallel encoder and dynamic attention fusion module, the dynamic coupling relationship between ECG and sleep signals is quantified to generate a sudden cardiac death risk score. This solves the problems of low early warning accuracy and device complexity in existing technologies, and enables precise risk monitoring in a home environment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 王伟玲
- Filing Date
- 2026-06-10
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies rely solely on single-modal signals or fail to effectively analyze the dynamic coupling relationship between multimodal physiological signals, resulting in low accuracy in predicting the risk of sudden cardiac death. Furthermore, traditional devices are bulky, complex to operate, and expensive, making it difficult to conduct routine, long-term monitoring in a home environment.
By synchronously acquiring dynamic electrocardiogram (ECG) signals and multimodal sleep physiological signals, feature sequences are extracted using parallel current encoders and sleep physiological flow encoders. Combined with a cross-modal dynamic attention fusion module and a temporal context aggregation module, the dynamic coupling relationship between ECG and sleep signals is quantified, a probability score for the risk of sudden cardiac death is generated, and interpretable output is provided.
It enables accurate and interpretable early warning of the risk of sudden cardiac death, improves the accuracy of early warning, simplifies the wearing and operation of the device, reduces costs, and is suitable for long-term home monitoring.
Smart Images
Figure FT_1 
Figure FT_2 
Figure FT_3