Table 3.

Feature additions of the ASF-24P stroke-oriented multidomain feature set.

Domain	Feature	Internal Parameters	Short Description
Time domain	MMAV2,MMAV5	$w_n=\left\{\begin{array}{c}1, 0.25N\le n\le 0.75\mathrm{N}\\ \frac{4n}{N}, n<0.25N\\ \frac{4\left(N-n\right)}{N}, otherwise\end{array}\right.$; $w_n=\left\{\begin{array}{c} \frac{4n}{N}, n<0.25\mathrm{N} \\ \frac{4n}{N}-1, 0.25N\le n\le 0.5\mathrm{N}\\ \frac{4n}{N}-2, 0.5N\le n\le 0.75\mathrm{N}\\ \frac{4n}{N}-3, otherwise\end{array}\right.$;	$\mathrm{MMAV}2=\frac{1}{N}{\displaystyle \sum }_{n=1}^N{w}_n\left|x_n\right|$; $\mathrm{MMAV}5\mathrm{splits}\mathrm{the}\mathrm{signal}\text{’}\mathrm{s}\mathrm{energy}\mathrm{window}(w_n$) of interest aiming to investigate the 3/5th of 4/5th window segment. Equitation is similar to MMAV2 [25,57].
	SSI		$\mathrm{SSI}={\displaystyle \sum }_{n=1}^N{\left|x_n\right|}^2$;
	KURT		$\mathrm{KT}=\frac{N\left(N+1\right)}{\left(N-1\right)\left(N-2\right)\left(N-3\right)}{\displaystyle \sum }_{n=1}^N{\left(x_n-\overline{x}\right)}^4-3\frac{{\left(N-1\right)}^2}{\left(N-2\right)\left(N-3\right)}$;
	SD		$\mathrm{SD}=\sqrt{\frac{1}{N}{\displaystyle \sum }_{n=1}{\left(x_n-\overline{x}\right)}^2}$;
	MFL		$\mathrm{MFL}={log}_{10}\left(\sqrt{{\displaystyle \sum }_{n=1}^N{\left(x_n-x_{n+1}\right)}^2}\right)$;
	WL	Threshold set to 0.05	$\mathrm{WL}={\displaystyle \sum }_{n=1}^{N-1}\left|x_{n+1}-x_n\right|$;
	MHW		$\mathrm{MHW}={\displaystyle \sum }_{n=0}^N{\left(W_n{x}_n\right)}^2$;
	AR3	$x_n=a_0+a{r}_1\left(x_n-1\right)+a{r}_2\left(x_n-2\right)+a{r}_3\left(x_n-3\right)$;	$\mathrm{AR}3=\left[a_0,a{r}_1,a{r}_2,a{r}_3\right]$, order set to 3
	LPC3	$x_n=b_0+b_1{x}_n+b_2(x_n-1)+b_3\left(x_n-2\right)$;	$\mathrm{LPC}3=\left[b_0,b_1,b_2,b_3\right]$, order set to 3
Frequency domain	MASP	$\mathrm{Fast}\mathrm{Fourier}\mathrm{transform}\mathrm{parameters}\mathrm{are}\mathrm{split}\mathrm{into}5\mathrm{bins}ff{t}_k\left(k=1, 2, 3, 4, 5\right)$	$\mathrm{MASP}={\displaystyle \sum }_{n=1}^N\frac{\left|ff{t}_k\right|}{N}$;
	SMN		$\mathrm{SMN}=\frac{{{\displaystyle \sum }}_{j=1}^M{f}_{j}PS{D}_j}{{{\displaystyle \sum }}_{j=1}^{M}PS{D}_j}$;
	MMNF		$\mathrm{MMNF}=\frac{{{\displaystyle \sum }}_{j=1}^M{f}_j{A}_j}{{{\displaystyle \sum }}_{j=1}^M{A}_j}$;
Time–frequency domain	STFT		$\mathrm{SFTT}={\displaystyle \sum }_{r=1}^{N-1}x\left(r\right)g\left(r-n\right){ϵ}^{-j2\pi mi/N}$;
	EWT	Wavelet Daubechies (Db4)	$\mathrm{EWT}=\sqrt{\frac{1}{K}{\displaystyle \sum }_{k=1}^K{W}_{J,k}^2}$;
	STW	STW is a noise-resilience method of wavelet transform and STFT to highlight signal window length other than artefacts or defect stochastic window frames [60].
Fractal domain	HFD	HFD evaluates muscle strength and the contraction grade; it measures the size and complexity of the sEMG signal in the time-domain spectrum without fractal attractor reconstruction methods [58].

g is a window function (STFT); $f_j$ is a frequency of the signal amplitude spectrum (MMNF, SMN); $W_n$ is a Hamming window function (MHW); K is the number of the j-th layer decomposed coefficient level, and $W_{J,k}$ is the k-th coefficient of the given layer decomposed coefficients (EWT).