A novel approach to diagnose sleep apnea using enhanced frequency extraction network

•We propose an enhanced frequency extraction network. It is an optimal architecture that automatically extracts sleep apnea-hypopnea events features from the spectrogram of nasal airflow signals.•The designed module in the proposed network to meet the specific demands in sleep apnea-hypopnea events detection.•In the domain of detecting sleep apnea-hypopnea events, the proposed model achieves superior performance as compared to previous methods using machine learning and deep learning