Abstract
The retina is susceptible to numerous diseases, and cataracts are most prevalent, especially in developing nations. Cataracts are recognized as one of the most impactful diseases affecting the retina, given their propensity to develop asymptomatically and potentially lead to blindness or impaired vision among the elderly. Timely detection of cataracts and appropriate intervention is pivotal in mitigating disease progression and reducing instances of blindness attributable to this condition. This study provides a deep learning system based on parallel architectures, that utilized a proposed deep convolutional neural network (DCNN), to detect and diagnose cataract disease accurately. ODIR dataset was used for training and validating the proposed model, which achieved 97.7% accuracy for cataract detection, with an inference time of no more than 0.06 sec.