Algorithm 1: Pseudocode of proposed framework

Input: OCT images

Output: Results

01: Load OCT image dataset

02: for k = 1 to 1000 do

03:    Read each image

04:    for j = 1 to 18 do//Feature generation using 18 pre-trained networks

05:      $X^j\left(k,1:1000\right)=CN{N}^j\left(Image\right);$//Extract deep features using jth CNN

06:      $cnt=1000$;//Counter defining to calculate the number of features.

07:      for i = 1 to 3 do//Creating multilevel feature generation network

08:         $c^1=max{p}^1\left(Image,\left[3\times 3\right]\right);$//Apply maximum pooling with 3 × 3 sized blocks

09:       $c^2=max{p}^2\left(Image,\left[3\times 3\right]\right);$ //Apply max-mean pooling

10:       $c^3=max{p}^3\left(Image,\left[3\times 3\right]\right);$//Apply max-min pooling

11:               $X^j\left(k,cnt+1:cnt+3000\right)=conc\left(CN{N}^j\left(c^1\right),CN{N}^j\left(c^2\right),CN{N}^j\left(c^3\right)\right);$

//In Line 11, $conc(.)$ defines concatenation operator and pre-trained CNN generates 3000 features from compressed images.

12:       $cnt=cn+3000$

13:       $new=max{p}^1\left(Image,\left[2\times 2\right]\right);$//Compress using images

14:       $I=new;$

15:      end for i

16:    end for j

17: end for k

18: for j = 1 to 18 do

19:    Select the best 1000 features ($f^j$) from $X^j$ with a length of 10,000.

20:    Calculate loss values deploying SVM classifier with 5-fold cross-validation

21: end for j

22: Select the best five features using calculated loss values. We have used quadratic support vector machine (QSVM) as a loss value generator in this phase. An error array with a length of 18 is created using this classifier. The optimal five CNNs are chosen using the created loss array. The minimum loss valued CNNs is the optimal performing CNNs.

23: Concatenate these features and obtain 5000 sized feature vector.

24: Apply IRF to 5000 sized feature vector for selecting the best feature vector.25: Classify the selected feature vector using SVM and obtain predicted results.