Diabetic Retinopathy Grading Using Multi-Scale Residual Network with Grouped Channel Attention

Diabetic retinopathy(DR) is an eye condition that can cause vision loss and blindness in people who have diabetes. It affects blood vessels in the retina which is the light-sensitive layer of tissue in the back of our eye. Accurate classification of diabetic retinopathy is crucial for early intervention and effective management. In recent years, deep-learning models have been employed to classify diabetic retinopathy based on retinal images. However, challenges such as small and imbalanced datasets can impact model performance. In this paper, we explore the effectiveness of Res2Net with Grouped Channel attention (GCA) for diabetic retinopathy classification. The proposed framework captures intricate patterns and relevant information, leading to better performance in identifying retinal abnormalities.