MCFN: A Multichannel Fusion Network for Sleep Apnea Syndrome Detection

Table 3

Performance comparison between MCFN and existing methods.

	Signals	Methods	Patients	Classify	Accuracy (%)
Gutiérrez et al. [15]	Flow	AdaBoost	317	Apnea/normal	86.5
Lin et al. [13]	Flow, Abdo. Thor.	SVM	34	OSA/CSA/hypopnea	81.8
Jiménez et al. [16]	Flow, SpO2	AdaBoost	974	OSA/normal	81.3
Haidar et al. [23]	Flow, Abdo. Thor.	CNN	2056	OSA/hypopnea/normal	83.4
Van Steenkiste et al. [26]	Abdo. Thor. EDR	LSTM	2100	Apnea/normal	77.2
Elmoaqet et al. [27]	Flow, Abdo. NPRE	LSTM/Bi-LSTM	17	OSA/CSA/MSA	83.6
Barroso et al. [31]	Flow, ODI3	MLP	946	Apnea/normal	82.5
Yu et al. [32]	EEG, flow, Abdo.	LSTM_CNN	126	Normal/hypopnea/OSA/MSA	83.9
Ours	Flow, Abdo. Thor.	MCFN	2056	OSA/Hypopnea/normal	87.3