From: Brain–computer interface robotics for hand rehabilitation after stroke: a systematic review
Component | Features and specifications |
---|---|
Data acquisition system and software | Dry EEG system with 8–16 channels, comfortable and easy to use Inclusion of other bio-signal sensors such as EMG, EOG, force, accelerometers to remove artifacts and improve classification Robust and reliable signal processing software: machine learning-based algorithms that discriminate brain states such as MI or evoked potentials with high classification accuracies (≥ 95%) and lower calibration times |
Hand Robot | Safe, comfortable and aligned with the hand’s range of motion Effective in providing kinaesthetic feedback Use of back-drivable or soft actuators that effectively assist movement without additional injury Multiple levels of safety and emergency features (mechanical, electronic, software), clear and obvious operation |
Visual cue and feedback | Provide rich visual cue and feedback to intended tasks, geometric representation of the hand (video or simulated environment), can be in multiple platforms such as display monitors or VR/AR on a head-mounted device Gamification of therapy exercises to provide an engaging regime to stroke patients |