Atrial fibrillation classification using deep learning algorithm in Internet of Things-based smart healthcare system

Detecting the electrocardiogram pattern in Internet of Things-based healthcare system and notifying this to the user is a challenging task. Using advance computing methods for classification of electrocardiogram signal is a notable research topic. In this research work, an intelligent electrocardiogram signal classification, employing deep learning algorithm, developed and tested in Internet of Things-based smart healthcare system was proposed. For classification of acquired electrocardiogram signal, a partitioned deep convolutional neural network was proposed. The electrocardiogram feature continuously in the Internet of Things-based monitoring system was learnt. To make use of learned features in the continuous time series data, it forms a higher order space in the server. We have made quantifiable comparative analysis with other classification algorithm with the same time series data collected from different atrial fibrillation samples in the Internet of Things-based e-health system. Our proposed algorithm learned features were tested in atrial fibrillation classified signal with other conventional classifiers with various performance indices. We obtained an accuracy of 96.3 percent with 93.5-percent sensitivity and 97.5-percent precision. From the obtained result, processing with proposed deep convolutional neural network provides reliable timely assist and accurate classification of electrocardiogram signal in Internet of Things-based smart healthcare system.