This item is available under a Creative Commons License for non-commercial use only
Computer Sciences, *human – machine relations
Fall prevention for geriatric populations is a growing concern among clinicians and researchers due to severe risk of morbidity and loss of independence. Emerging evidence has demonstrated that mental workload while walking influences gait stability and the risk for falling. Electroencephalography (EEG) presents a potential method to provide objective measures of mental workload, particularly during daily activities. Noise introduced to the EEG signal during motion, however, is restrictive. The study presented in the following paper isolates EEG signal noise attained from gait for a commercially accessible EEG system, the "Emotiv" Time and spectral system identification techniques were applied to model motion-induced artifacts given head linear and angular movement data. During gait of varying speeds (1.4-5.8 km/h), frequency and time domain system identification techniques were unable to accurately model the relationship between head movement and EEG signal noise with accuracies between 1% and 11% fit. However, analysis of data obtained during activities eliciting higher amplitudes of head movement (i.e., double-footed jumping) resulted in a high accuracy in linear system modelling, ranging from 68% to 74% suggesting a dead-zone non linearity. Isolated EEG noise signals provide a ground truth measurement of the "Emotiv" system to estimate signal to noise ratio (SNR). Results of the SNR determined that, during gait, EEG signals are 8 to 20 times the power of gait-induced noise providing confidence in EEG recordings during ambulatory monitoring.
Shea, K. & Tung, J. (2017). System identification of motion artifact noise in EGG headsets from locomotion. H-Workload 2017: The first international symposium on human mental workload, Dublin Institute of Technology, Dublin, Ireland, June 28-30. doi:10.21427/D75349