Deep Learning Methods for the Prediction of Chronic Diseases: A Systematic Review

Deep learning (DL) is a machine learning optimization technique that has represented amazing performance in identifying obscure structures in high-dimensional data to find the most optimal settings. Thus, DL has been effectively applied in many diverse fields in image and speech recognition, visual art, natural language processing, and bioinformatics. Other than this, lots more are still needed to be investigated. This paper systematically reviews publications used in deep learning methods for the prediction of chronic diseases more accurately. In medical science, it is always challenging to analyze chronic diseases before the major damages. Some of the chronic diseases cannot be recognized in primary diagnosis until they put a drastic impact on health, as some of them have no treatment. Hence, to avoid such an awful condition, there is a sturdy requirement of some models that can predict disease more accurately in an early stage. Different models have been designed using deep learning’s multilayer approach and provide better result in the prediction of some chronic diseases that comprises Coronary Heart disease, Alzheimer disease, labeling of multiple chronic disease, Diabetic Retinopathy, Breast cancer, Autoimmune disease, and skin diseases. This paper summarizes all of these DL models for predicting the mentioned diseases.