Algorithmic Analysis of Distance-based Monarch Butterfly oriented Deep Belief Network for Diabetic Retinopathy

This paper proposes a newDiabetic Retinopathy detection model with four phases including (i) Preprocessing, (ii) Blood Vessel Segmentation, (iii) Feature Extractionand (iv) Classification. At first, pre-processing is performed using the Contrast Limited AHE (CLAHE) and median filtering methods. To segment the blood vessels, Fuzzy C-Mean (FCM) thresholding is worked well for making rough clustering of pixels. Further, features like local features and morphological transformation-based features are extracted from the segmented blood vessels. Finally, Deep Belief Network (DBN) classifier is used to classify the extracted features into a healthy and affected image. Optimization logic is undertaken in this work, where the number of hidden neurons in DBN is optimized using modified Monarch Butterfly Optimization (MBO) termed as Distance-based MBO (D-MBO). Finally, Algorithmic analysisis performed under different performance measures like positive and negative measures: Accuracy, Sensitivity, Specificity, Precision, FPR, FNR, NPV, FDR, F1 Score and MCC.