[Retracted] Automatic Detection of High-Frequency Oscillations Based on an End-to-End Bi-Branch Neural Network and Clinical Cross-Validation

Table 6

The comparison between proposed method and existing methods using intrasubject validation.

Methods	SEN (%)	SPE (%)	PRE (%)	ACC (%)	FDR (%)	F1 (%)	SEN_SPE (%)
[44]	81.1 (R)	—	—	—	30.2 (R)	—	—
	74.6 (FR)				6.3 (FR)
[21]	91.26	91.52	88.67	—	—	89.95	91.39
[27]	87.40	77.60	—	—	—	—	82.21
[45]	80.9	96.9	—	90.7	—	—	88.18
[46]	100	33	—	78	—	—	49.62
Proposed	94.62	92.70	92.12	93.62	7.88	93.33	93.63