Heartbeat detection in ballistocardiograph using a deep learning based neural network
Date of Issue2018
School of Electrical and Electronic Engineering
Ballistocardiography (BCG) is a revamped technology for cardiac function monitoring. Detecting individual heart beats in BCG remains a challenging task due to various artifacts and low signal-to-noise ratio, which are not well addressed by conventional approaches. In the dissertation, we propose to employ deep learning networks to capture the distinguishing characteristics of various BCG waveforms within and across individual subjects. Particularly, we design a neural network that combines Convolutional-Neural- Network (CNN) and Extreme Learning Machine (ELM) together. In order to verify the effectiveness of our proposed method, we construct a signal acquisition system and collect simultaneous ECG and BCG signals with high quality from healthy adult volunteers. We examine the new learning method on the new dataset as well as on an existing BCGECG dataset. The result shows a significantly higher detection accuracy by the proposed method than a state-of-the-art method. We demonstrate how the advanced machine learning technology can learn and detect BCG waveforms robustly. Parts of this work have been accepted for publication in IEEE Engineering in Medicine and Biology Society (EMBS) 2018 conference.
DRNTU::Engineering::Electrical and electronic engineering