Deep Transfer Learning for Visually Induced Motion Sickness Detection Using Symmetric Projection Attractor Reconstruction of the Electrocardiogram

Despite the ubiquity of motion sickness – a long recognized syndrome from ancient days of sea travel – it still remains a persistent problem. In fact, around one in three individuals can be severely susceptible to this malady. The electrocardiogram (ECG) is an essential tool that has long been used to examine the physiological expression of motion sickness; commonly by performing analysis of ECG-derived heart rate variability (HRV). Here, we obtained the symmetric projection attractor reconstruction (SPAR) transforms of ECG signals recorded from healthy participants at rest and during nausea, for a binary image classification task using a set of pretrained deep neural networks with transfer learning. Our observations provide new insights into how physiologic characteristics captured via ECG-derived attractor images may be important for the detection of ECG signals that show differential response to motion-induced nausea.