MSE-VGG: A Novel Deep Learning Approach Based on EEG for Rapid Ischemic Stroke Detection

. 2024 Jun 29;24(13):4234. doi: 10.3390/s24134234

Algorithm 1 Algorithm for feature extraction

Input: EEG database $N = N_{n}^{s}$ , where n is the number of EEG channels, s means the number of subjects.
Output: RGB map of cwPLI+SaEn
for $p = 1$ to s do
preprocess the EEG signal
dataset $D = D^{d} \leftarrow$ segmented each $N_{n}^{p}$ into epochs of 1 s
${PLI}^{p} \leftarrow$ calculate the PLI matrix of each $N_{n}^{p}$
for $i = 1$ to n do
for $j = 1$ to n do
$w_{i j} \leftarrow$ calculate the correlation weight of each $P L I_{i j}$ in ${PLI}^{p}$ matrix
end for
end for
$w \leftarrow w_{i j}$
end for
for $m = 1$ to d do
${PLI}^{m} \leftarrow$ calculate the PLI of each $D^{m}$
${cwPLI}^{m} \leftarrow {PLI}^{m} \times w$
$S a E n^{m} \leftarrow$ calculate the SaEn of each $D^{m}$
( $cwPLI + SaEn$ ) $^{m} \leftarrow {PLI}^{m} + S a E n^{m}$
end for
Convert $cwPLI + SaEn$ to RGB map
return Outputs