Deep Learning Frameworks in Healthcare Systems

Abstract

Consider we are living in a remote place which is far away from any near-by hospitals or we don’t have enough time to take a leave from the work to visit a hospital or we can’t afford the rapidly increasing medical services costs. This is one of the few scenarios where we can relate to the fact that computer systems and algorithms if they may help in detection or diagnosis of the diseases can be revolutionary and life changing. Over the last decade there has been an increase in the study of how computers, particularly through the disruptive technologies like deep learning, can help in the pervasive sensing for health and well-being. Several researchers have proposed numerous works to detect diseases such as Cancer, RA, heart or lung diseases, Diabetic Retinopathy, Parkinson’s Disease, etc. Deep learning techniques enable computers to adapt and self-learn from the information fed into the systems by continuously recognizing the structures and patterns in them. This helps the computers to detect and diagnose a particular disease. This paper focuses on the recent studies in the field of deep learning which can be used for the disease detection and diagnosis. They have been discussed in correlation with the benefits of deep learning in the healthcare sector and their future scope in the same. We have proposed three case studies to highlight the significance of deep learning in healthcare sector by highlighting relevant use cases in each of them. First case study assess the age-related macular degeneration from fundus images by using DNN screening technique. Second case study uses CNN model for the exudate detection of diabetic retinopathy. Third case study uses Deep learning technique to classify cataract fundus images. Finally, the observations made in each of the case studies have been compared and analyzed.