Diabetic Retinopathy (DR) is a chronic eye complication caused by a metabolic disease that affects an individual’s vision. At first, symptoms are not visible, but if left undiagnosed, they may lead to vision impairment. In this chapter, a framework is proposed to detect diabetic retinopathy, and a health monitoring system is designed that ensures long-term care of patients suffering from diabetic retinopathy. One major obstacle faced by the medical domain is the lack of huge labeled data, hence real-world applications demand solutions that require less labeled data while still be able to deliver impactful outcomes. Therefore, it is necessary to develop an explainable self-supervised diabetic retinopathy severity detection model that utilizes a pre-trained model trained on large unlabeled fundus images to perform the end task where labeled data is scarce. This work proposes an integrated model for predicting the severity of diabetic retinopathy (DR) by incorporating data from electronic medical records (EMR) and mobile application. This integration is anticipated to play a crucial role in delivering comprehensive care to patients suffering from DR.