SE-MSCNN: A Lightweight Multi-scaled Fusion Network for Sleep Apnea Detection Using Single-Lead ECG Signals
Abstract
Sleep apnea (SA) is a common sleep disorder that occurs during sleep and its symptom is the reduction or disappearance of respiratory airflow caused by upper airway collapse. The SA would cause a variety of diseases like diabetes, chronic kidney disease, depression, cardiovascular diseases, or even sudden death. Early detecting SA and intervention can help individuals to prevent malignant events induced by SA. In this study, we propose a multi-scaled fusion network named SE-MSCNN for SA detection based on single-lead ECG signals acquired from wearable devices. The proposed SE-MSCNN mainly has two modules: multi-scaled convolutional neural network (CNN) module and channel-wise attention module. To utilize adjacent ECG segments information to facilitate the SA detection performance, the multi-scaled CNN module consists of three streams