Early detection and prediction of Heart Disease using Wearable devices and Deep Learning algorithms

In this paper, we propose a multimodal deep learning algorithm that combines convolutional neural networks (CNNs) and long short-term memory (LSTM) networks for early detection and prediction of heart disease using data collected from wearable devices. This combined multi-model deep learning algorithm is used to detect the accurate precision and accuracy value. At first, we consider, ECG and PPG signals, which are collected from the dataset. Then, the features from ECG and PPG are extracted using CNN and the accelerometer features are extracted using the LSTM model. The combined features are then classified using hybrid CNN-LSTM network architecture. The algorithm is evaluated using a publicly available benchmark dataset. The model achieved an accuracy of 99.33% in detecting heart disease, outperforming several state-of-the-art deep learning models. In addition, the model can predict the likelihood of developing heart disease with a precision of 99.33%, providing an early warning system for at-risk patients. The results demonstrate the potential of a multimodal approach for early detection and prediction of heart disease using wearable devices and deep learning algorithms.