Algorithm 1 Multiscale entropy (MSE) computation

For each EEG channel, define a TS as $X=\left[X\right(1),X(2),\cdots ,X(\ell \left)\right]$ where ℓ is the number of EEG samples. Fix m, the length of compared sub vectors Choose ${\tau }_{min}$, $\Delta \tau$, ${\tau }_{max}$, the minimum, maximum and increment values, resp., of the subsample vector $\tau$. for $\tau$ = ${\tau }_{min}$ to ${\tau }_{max}$ step $\Delta \tau$ do  Subsample X for ${\tau }_{min},\tau \left(2\right),\dots ,{\tau }_{max}$  Fix q sub TS from each EEG channel as $X_1=\left[X\left(1\right),X\left(2\right),\cdots ,X\left(\frac{\ell }{q}\right))\right]$,…, $X_q=[X(\ell -q),\cdots ,X(\ell )]$.  for i=1 to q do    Determine a variable threshold $r\left(X_i\right)$ for each sub TS, $X_i$ according to $r\left(X_i\right))=0.2\sigma \left(X_i\right)$    Construct a sequence of m-long vectors as $z=\left\{u\right(1)$, $u\left(2\right)$, …, $u(\ell -m+1)\}$ as $u\left(i\right)=\left[X\right(i\left)X\right(i+1\left)X\right(i+1-m\left)\right]$    From the sequence z, calculate the number of $u\left(j\right),j=N/\tau$ such that its distance $d(u\left(i\right),u\left(j\right)))=ma{x}_i|u\left(i\right)-u\left(j\right)|<r\left(X_i\right)$    For $i\ne j$, calculate the number of distances (denoted as #) within threshold $r\left(X\right(i\left)\right)$ as $C_i^m\left(r\right)=\frac{\#d\left(u\right(i),u(j\left)\right))<r}{\ell -m+1}$    Calculate the logarithmic average ${\Phi }^m\left(r\right)=\frac{1}{\ell -m+1}{\sum }_{i=1}^{\ell -m+1}$ln$\left(C_i^m\left(r\right)\right)$    For each sub TS vector, compute $S_i={\Phi }^m\left(r\right)-{\Phi }^{m+1}\left(r\right)$ = $\frac{1}{\ell -m+1}{\sum }_{i=1}^{\ell -m+1}$ln$\left(\frac{C_i^m\left(r\right)}{C_i^{m+1}\left(r\right)}\right)$    Determine the average of all $S_i$ and name it $MS{E}_i$  end for end for Plot $\tau$ vs. MSE