Enhancing CNNs through the use of hand-crafted features in automated fundus image classification

Eye diseases such as diabetic retinopathy and macular edema pose a major threat in today’s world they affect significant portion of the global population. Therefore, it is utmost importance to develop robust solutions that can accurately detect these diseases, especially their early stages. However, current methods, based on hand-crafted features devised by experts, are not sufficiently accurate. Several have been proposed use deep learning techniques improve performance systems. ignore highly valuable features, could contribute more accurate prediction, which underlines significance our research. In this paper, we revisit problem combining with extracted neural networks objective delivering predictions. We systematically study several state-of-the-art methods propose number ways integrate them into framework. show arrived at conclusion possible achieve significantly better results outperform do consider using methods.