Abstract
Diabetic retinopathy (DR) is the main cause of blindness in diabetic patients.
Early and accurate diagnosis can improve the analysis and prognosis of the disease.
One of the earliest symptoms of DR is hemorrhages in the retina. Therefore, we
propose a new method for accurate hemorrhage detection from retinal fundus images.
Here, the proposed method uses the modified contrast enhancement method to improve
the edge details from the input retinal fundus images. In the second stage, a
convolutional neural network (CNN) with improved LSTM based on hybrid Harris
Hawks with Mayfly (HHMO) is proposed to detect and classify the hemorrhages.
Finally, the proposed CNN with HHO-LSTM is compared with the existing techniques
including machine learning and deep learning techniques such as Naïve Bayes, SVM,
ANN, etc., and traditional CNN, LSTM, and other techniques, respectively. Therefore,
the comparison can prove that the proposed model is more effective in detecting and
classifying Hemorrhages in the retina due to diabetic retinopathy. The performance
metrics considered in this work are accuracy, specificity, sensitivity, f1-score,
precision, etc.