A machine learning-enabled intelligent application for public health and safety

This article uses a machine learning approach to analyze and detect burst signals in a smart healthcare system to protect and safeguard the public health. We consider the time-series electrocardiogram (ECG) waveforms for the detection of burst signals. For this purpose, we propose an intelligent differential correlation burst detection (DCBD) approach by establishing a mathematical model and deriving the analytical expressions to detect false alarm rate and missed detection rate in ECG signals. DCBD feeds the burst signals of a large-scale ECG waveform into various filters for noise removal, which are then passed through data augmentation to achieve high specificity and sensitivity. These waveforms are then segmented for feature extraction and machine learning (ML) classification. Finally, burst-free ECG waveforms are broadcast to a database server, where ML algorithms are used to detect the presence of any abnormal activities. Furthermore, the ECG signal is classified to a set of heart diseases using the well-known LSTM (Long Short-Term Memory) and CNN (Convolutional Neural Network) models. Our proposed approach highlights that the probability of false alarm rate is similar to that caused by pure noise within the ECG waveforms. Our evaluation, using numerical experiments, suggests that the accuracy of the LSTM based ECG signal classification could be approximately 11.7% and 12.8% improved, subsequently, using the proposed burst detection method.