Sleep apnea (SA) is a common sleep-related breathing disorder and it remains a challenge to effectively and accurately detect SA out of hospital. Assistant Prof. Fan Xiaomao from College of Big Data and Internet at Shenzhen Technology University (SZTU) and his team members focus on SA detection with single lead ECG signals, which can be easily collected by a portable device. Under this context, a restricted attention fusion network called RAFNet for sleep apnea detection was proposed. Extensive experiments show that RAFNet can achieve 91.40% accuracy on Apnea-ECG and 84.70% accuracy on FAH-ECG. It can also obtain 100% on SA detection performance metrics of accuracy, sensitivity, and specificity as well as 85.56% per-recording accuracy. It means that RAFNet can be potentially deployed into a medical system to provide sleep conditions monitoring service for a large number of population.

The architecture of RAFNet [Photo/https://www.sciencedirect.com/science/article/pii/S0893608023001454]

The research team published an article titled “RAFNet: Restricted attention fusion network for sleep apnea detection” in Neural Networks (IF: 9.657). The research was jointly conducted by SZTU researchers and Prof. Lei Wenbin and his team from the First Affiliated Hospital of Sun Yat-sen University.

The full text of the article can be found at: https://www.sciencedirect.com/science/article/pii/S0893608023001454.

Drafted by Daisy（姚琦）/ International Cooperation & Student Affairs Office

Revised by International Cooperation & Student Affairs Office

Edited by International Cooperation & Student Affairs Office