Diabetic Retinopathy Prediction Using DenseNet 201 with Swish as Activation Function

Diabetes Retinopathy (DR) is one of the important complications of diabetes, DR emerges as a serious complication associated with diabetes mellitus causing damage retinal blood vessels and resulting in vision loss. This study used DenseNet121 and DenseNetl69, DenseNet201, EfficientNetBO, EfficientNetB7, Inception ResNetV2, Inception V3, MobileNet, MobileNetV2, MobileNetV3Large, ResNet50, ResNetl0l, and ResNet152 13 different deep learning neural network models were used for DR prediction. The change of Validation Accuracy trend was initially observed with 10,20,30 and 40 different EPOCHs. Before grading, pre-processing such as elimination, amplification, denoising, data enhancement, normalization, etc. were carried out for the problems in the retinal image data set in the experiment. In order to avoid vanishing gradients in the neural network model and ensure classification accuracy, DenseNet201 was adjusted by using a targeted swish function for processing, and classify retinal images into three levels, and finally the classification accuracy of the neural network model proposed in this study reached 99.22%, macro avg precision 99.22%, macro avg recall 99.220/0, macro avg f1-score 99.22%, It is significantly better than other models, and on average, it can correctly identify and automatically grading more than 6,870 images In the verification of 6,926 retinal images. The method proposed in this study has been validated to effectively improve the accuracy of DR classification and can be used for large-scale DR medical imaging screening, effectively assisting doctors in achieving efficient diagnosis of DR grading.