Automatic Sleep Stage Classification Based on Convolutional Neural Network and Fine-Grained Segments

Table 6

Evaluation results of our proposed method succeeding 30 s.

Networks	Stages	Precision		Sensitivity		Specificity		Accuracy
		Best	Worst	Best	Worst	Best	Worst	Best	Worst
Proposed method	W	0.9200	0.8351	0.9716	0.7612	0.9876	0.8700	0.9742	0.8078
	N1	0.8790	0.6384	0.8674	0.69	0.9744	0.7812	0.9433	0.8422
	N2	0.9417	0.6054	0.9328	0.5079	1	0.8054	0.9645	0.8023
	N3	0.9512	0.7028	0.9766	0.6954	1	0.8912	0.9812	0.9065
	REM	0.9347	0.7700	0.8906	0.7822	0.9972	0.8433	0.9365	0.7612
	Average	0.9274		0.9253		0.9477		0.9220