Review of Deep Learning for Physiological Signal-based Healthcare Applications

Abstract

Deep Learning has been considered as the successful approach for generated as well as discriminative task which is recently proved with medical analysis of two-dimensional images and it is highly potential even though the physiological data are in the form of signals of single dimension the beneficially should be analyzed using the novel approach of various healthcare applications based on the physiological signals using deep learning. The studies include recent scientific researchers done on deep learning based on the physiological signal data that includes the electrocardiogram (ECG), electromyogram (EMG), electrooculogram and electroencephalogram. The main goal of the research is the study of deep learning for the physiological signal-based model for various healthcare applications and development of physiological signal-based model for stress detection in working people and wearable physiological system. It helps to compare, categorize and comprehend the parameters off deep learning approaches for the analysis of physiological signals. Dataset resources, deep learning model, type of input data and healthcare applications are compared. In the perspective of signal data analyzing the physiology for healthcare applications and data modality as well as the concept of system CNN architecture based on deep learning and using data set.