Table 3.

The comparison results of the normalized prediction N-RMSE values among different studies

Approaches	HNM (Zhang et al., 2022b)	SVR (Zhang et al., 2022a)	FFNN (Zhang et al., 2022a)
N-RMSE [%]	9.48  1.51	8.49  2.74	10.13  3.44
	0.91  0.04	0.93  0.05	0.89  0.08
Approaches	CNN–ROI1	CNN–ROI2	CNN–ROI3
N-RMSE [%]	8.97  1.33	7.34  1.09	6.91  0.95
	0.89  0.03	0.93  0.03	0.94  0.02

Note. Reported data are the mean ± standard deviation across conditions and participants, where previous studies applied data fusion between sEMG and US imaging signals.

Abbreviation: HNM, hill-type neuromuscular mode; SVR, support vector regression; FFNN, feedforward neural network.