Table 5.

Frequency-domain features and their respective equations that were considered in this work.

Time-Domain feature	Equation
Spectral entropy	$H=-{\sum }_i{p}_i·lo{g}_2(p_i),$ with $p_i=\frac{P_i}{\sum _{i=0}^{L-1}{P}_i}$
Mean spectral power	$\overline{P}=\frac{1}{L}·\sum _{i=1}^L{P}_i$
Maximum spectral power	maxP = max(P)
Spectral power variance	$va{r}_P=\frac{1}{L}\sum _{i=1}^L{(P_i-\overline{P})}^2$
Band power	$BP=\sum _{i=f_1/\Delta f}^{f_2/\Delta f}{P}_i$, with $\Delta f=\frac{1}{t_w}$
Epileptogenicity index	$Ep{i}_{index}=\frac{P_{\beta }+P_{\gamma }}{P_{\theta }+P_{\alpha }}$

P is the power spectrum vector, obtained by the DFT, f₁ the lower band frequency, f₂ the upper band frequency of the band power, and t_w the duration of the sampled window in the time-domain.