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

Table 9

The range of gains and losses in various indicators.

Methods	SEN	SPE	PRE	ACC	FDR	F1
Loss^1	2.17%	3.33%	—	—	—	—
Improvement^2	—	—	3.22%	5.73%	3.22%	7.47%

¹Loss degree of SEN and SPE using bi-branch model compared with 2 single-branch models. ²Improvement degree of SEN and SPE using 2 single-branch models compared with the bi-branch model.