Motion sickness detection through deep learning techniques

Motion Sickness is a disorder caused by incoherent inputs arising from the visual system and the vestibular system. Motion Sickness is subdivided into different cases: sickness due to visual stimulation, sickness due to vestibular stimulation and less frequently sickness due to somatosensory stimulation. Individual susceptibility to Motion Sickness varies according to different factors such as age, gender, physical condition. The most common symptoms caused by this disorder are pallor, vomiting, nausea, dizziness, frequent sweating, excessive salivation. This thesis focuses on the analysis of physiological signals to automatically detect patterns characterizing Motion Sickness disorders. The Electrocardiogram signal has been preprocessed by applying various types of filters, to reduce possible noise and artefacts. Subsequently, the target signal has been analysed using the Wavelet analysis coupled with the application of a Convolutional Neural Network (CNN) that has been trained and validated. In this way, the created Deep Learning model aims to classify different patterns characterizing the disorder of motion sickness.