|
Algorithm 1. Pseudo code of the proposed scheme. |
|
Proposed Algorithm |
01 |
ImP = Load (PeN-CoVx) // Read PeN-CoVx image dataset
|
02 |
ImNorm = (ImP – min(ImP)) / (max(ImP) – min(ImP)) // Normalize images within range [0, 1]
|
03 |
ImSWC = CSW(ImNorm) // Data augmentation: 3-levels decomposition by stationary wavelet transformation
|
04 |
ImAug = DataAugmentation(ImSWC) // Performing random rotation, translation, and shear operation
|
05 |
for i = 1 to size-of ImAug do // Loop to extract relevant features for each image using transfer-learning model
|
06 |
fg1 = AlexNet(ImAug) // Extract 1000 features via AlexNet
|
07 |
fg2 = ResNet101(ImAug) // Extract 1000 features via ResNet101
|
08 |
fg3 = SqueezeNet(ImAug) // Extract 1000 features via SqueezeNet
|
09 |
for j = 1 to 3 do // Loop to merge extracted feature
|
10 |
X(i, 1000 × j + 1: 1000 × (j + 1)) = fgj // merge extracted features
|
11 |
end for loop // end loop to merge extracted feature
|
12 |
end for loop |
13 |
fs1 = iNCA(X, Y) // Determine optimal features via iterative Neighborhood Component Analysis
|
14 |
fs2 = iChi2(X, Y) // Determine optimal features via iterative Chi-square
|
15 |
fs3 = iMRMR(X, Y) // Determine optimal features via iterative Maximum Relevance Minimum Redundancy
|
16 |
PLk = CNN(fsk, Y, 3) // Predict clinical-state using optimal features by Convolutional Neural Network
|
17 |
PLk = LDA(fsk, Y, 3) // Predict clinical-state using optimal features by Linear Discriminant Analysis
|
18 |
PLk = SVM(fsk, Y, 3) // Predict clinical-state using optimal features by Support Vector Machine
|