In today’s world healthcare system is getting more importance because of our life style and food habits. According to the World Health Organization report, approximately 32% of death rate in worldwide is due to Cardio-vascular disease. Coronary heart disease, Cerebro vascular disease, Rheumatic heart disease etc. are comes under this cardio vascular diseases. In healthcare systems, past years many machine learning techniques are deployed for analysis and prediction. In order to enhance and improve prediction accuracy, now a days, many deep learning techniques are providing more promising results in pattern recognition which in turn improves prediction accuracy in the healthcare systems. There are three components in the deep learning pipeline for cardiac ultrasonic imaging, they are data collection and preparation, network selection, training and evaluation. This survey paper an attempt is made to explores various deep learning techniques namely Convolutional Neural Networks (CNN), Deep Neural Networks (DNN), and Recurrent Neural Networks(RNN) for early disease prediction. Also explored about different types of data sets and feature extraction techniques involved in this prediction process. The main contribution of this survey paper, for past one decade different deep learning techniques involved in predicting Cardio-vascular disease and comparison has been made based on their accuracy with their dataset involved and feature extraction techniques used.