用于暂停呼吸综合征的睡眠呼吸检测方法、系统及云平台
By using stretch-sensitive fiber optic sensors and wavelet denoising technology in sleep apnea monitoring devices, combined with the intelligent recognition model of the cloud platform, the shortcomings of existing devices in signal processing and individual adaptability are solved, achieving highly accurate sleep apnea detection and risk assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU MEDICAL UNIV
- Filing Date
- 2025-06-19
- Publication Date
- 2026-07-17
AI Technical Summary
Existing sleep apnea monitoring equipment and platform systems have shortcomings in the universality and individual adaptability of signal processing algorithms, resulting in high false negative rates and numerous misjudgments, failing to meet the accuracy requirements of preclinical screening and risk assessment.
A stretch-sensitive fiber optic sensor was used to acquire chest wall motion signals. The signals were then denoised using the Daubechies fourth-order wavelet function and Stein's unbiased risk estimation method. The sliding window method was used to detect respiratory peaks and troughs. Dynamic amplitude and time thresholds were set, and a hybrid neural network of support vector machine and LSTM-TCN on a cloud platform was used for sleep apnea event identification and risk assessment.
It improves the accuracy and individual adaptability of respiratory signal processing, reduces the false alarm and missed detection rates, and achieves highly sensitive identification and risk assessment of sleep apnea events, making it suitable for scenarios such as home care, elderly care, and telemedicine.
Smart Images
Figure CN120419911B_ABST