REMOV: EEG artifacts removal methods during Lokomat lower-limb rehabilitation

Abstract

Post-stroke rehabilitation is one of the major health-care challenges. Robotic-aided therapy, if coupled with adequate monitoring techniques, is able to provide task-specific highly-intensive repetitive treatments that may facilitate patients' motor recovery. The EEG is the best non-invasive brain imaging modality in terms of sensors lightness, noninvasiveness, and temporal resolution, however artifact contamination has always made it difficult for scientists to use it in combination with lower limb robotic-aided rehabilitation. In this work we present for the first time REMOV, a method that combines various routines for the removal of EEG artifacts during Hocoma-Lokomat lower-limb rehabilitation. REMOV includes various preprocessing, abnormal data removal, channels rejection, ocular artifacts rejection and fine-tuning steps. This study, although at its preliminar state, may help scientists to use the EEG as brain imaging technique during Lokomat rehabilitation, and will hopefully pave the way to further advancements on EEG artifacts removal.