Heartbeat monitoring with an mm-wave radar based on deep learning: a novel approach for training and classifying heterogeneous signals

Millimetre wave radar is an emerging technology that can monitor vital signs without contact. This unique feature is very suitable for some particular situations, such as burn patient monitoring. Currently, electrocardiogram (ECG) is still the most common approach for monitoring heart disease. Deep learning algorithms have already been applied to classifying ECG recordings and have achieved good diagnostic results. However, it is very rare to see deep learning-based heartbeat classification using radar signals. The reason is a lack of radar-based heart disease datasets, which are the most important part of training a Convolutional Neural Network (CNN). Specifically, the ECG recordings and radar signals are heterogeneous; thus, the ECG dataset cannot train the CNN for directly classifying the radar signals. In this paper, we propose a novel signal processing algorithm called the Common Features Extraction Method (CFEM) to extract the common features of ECG recordings and radar signals to train a CNN for radar heartbeat signal classification. By using CFEM, the ECG dataset is transferred to the radar field, which means that the core issue for training the CNN using radar signals has been solved. Practical experiments show that the CFEM-based CNN can classify heartbeat radar signals accurately.