Estimating the neural spike train from an unfused tetanic signal of low-threshold motor units using convolutive blind source separation

Table 2 Comparing the performance of CBSS and HWM to estimate spikes based on experimental signals (voluntary and evoked contractions)

Voluntary contractions		Evoked contractions		CBSS	HWM	Pairwise	p-value
Unfused tetani	Spikes	Unfused tetani	Spikes	RoA (%)	RoA (%)	difference
3	58			98.0 ± 3.4	96.3 ± 3.2	1.7 ± 3.3	#
		18	720	98.0 ± 2.8	96.5 ± 5.6	1.1 ± 4.5	0.24

CBSS convolutive blind source separation, HWM Haar wavelet method, RoA rate of agreement (between estimated and EMG reference spikes). # To few unfused tetani to do a statistical test

ISSN: 1475-925X