Training in realistic virtual environments: Impact on user performance in a motor imagery-based Brain-Computer-Interface
Abstract
A brain–computer interface (BCI) is a system that enables people to control an external device by means of their brain activity, without the need of performing muscular activity. BCI systems are normally first tested on a controlled environment before being used in a real, daily scenario. While this is due to security reasons, the conditions that BCI systems users will eventually face in their usual environment may affect their performance in an unforeseen way. In this paper, we try to bridge this gap by presenting a trained BCI user a virtual environment that includes realistic distracting stimuli and testing whether the complexity or the type of such stimuli affects user performance. 11 subjects navigated two virtual environments: a static park and the same one with visual and auditory stimuli simulating typical distractors from a real park. No significant differences were found when using a realistic environment; in other words, the presence of different distracting stimuli did not worsen user performance.
Full text
14/6/15& 1& Training in realistic virtual environments: Impact on user performance in a motor imagery-based Brain-Computer Interface Leando da Silva-Sauer, Luis ValeroAguayo, Francisco Velasco-Álvarez, Sergio Varona-Moya, Ricardo RonAngevín Depto. Tecnología Electrónica Depto. Personalidad, Evaluación y Tratamiento Psicológico Universidad de Málaga Málaga (Spain) TEC-2011-26395 IWANN-2015 Training in realistic virtual environments • BCI in experimental contexts • Objectives • Method • Participants • Procedure • Instruments • Signal processing • Navigation paradigm • Virtual environment • Results • Conclusions IWANN-2015
14/6/15& 2& BCI in experimental contexts IWANN-2015 • Component of EEG (8-13 Hz) • Event Related Desynchronization • Navigation with ERD into VR (car, weelchair) • Imagination Responses: • Hand movement • Cognitive relax BCI in experimental contexts IWANN-2015 • Experimentation context, virtual reality • Lab tasks, artificial, non-natural • Generalization to natural context • Training more realistic spaces • Virtual space with distracting situations • Breaks in real situations on weelchair
14/6/15& 3& Objectives IWANN-2015 • Studying the effects • Realistic VR in a park scenario • Distracting situations (visual, movements, lights, auditory, surprise) • Application interface (visual, auditory) Method: Participants IWANN-2015 • 11 naïve voluntary students • Minimum error rate <30% in calibration • Informed consent • Bonus points on grade subject
14/6/15& 4& Method: Design IWANN-2015 • Intra-group with repeated measurements • Phases counterbalanced between participants • A – B1 – B2 – C2 – C1 (six participants) • A – B2 – B1 – C1 – C2 (five participants) A Calibration and adaptation B1 Static Stimuli Visual Auditory Interface B2 Distracting Stimuli Visual Auditory Interface C1 Static Stimuli Auditory Interface C2 Distracting Stimuli Auditory Interface Method: Procedure IWANN-2015 • 3 days or sessions, 1 hour aprox • First, calibration, mental imagination right hand and relax, no feedback • Second, navigation task B1 or B2 • Third, navigation task C1 or C2
14/6/15& 5& Method: Procedure IWANN-2015 • Navigation through a virtual park, making 21 movement commands Method: Instruments IWANN-2015 • Soundproff 4x2.4m room • Stereoscopic overhead projector screen • Polarized glasses • Sound system 5.1 surround
14/6/15& 6& Method: Instruments IWANN-2015 • EEG bipolar channels • c3 and c4 positions (hand sensorimotor areas) • Biosignal g.BSAMP (Guger Technologies) • NI-USB-610 National Instruments amplifier and digitalization (128 Hz and 12 bit) Method: Navigation Paradigm IWANN-2015 • Non-Control Command: • Semi-transparent vertical blue in center screen • Control Command: • Circle with three parts of quadrants, three navigation commands (forward, right, left) • Bar computer each 62.5 ms with classifier • If correct right-hand imagination task 1 s, bar extended • Audio cues for each quadrants • In C1 and C2 only auditive cues
14/6/15& 7& Method: Virtual Environment IWANN-2015 • Virtual Environment simulating a static park with or without distracting stimuli • 21 commands of a forced path in the park • Taks and time process: • Auditive cue indicating command (forward, right, left) • 1 second later, interface comand • Selection of a command • Feedback visual and auditory (positive or negative) Method: Virtual Environment IWANN-2015 • Virtual Environment with distracting stimuli while navigation and commands • Skater and boxes falling • Rain • Wind • Person reading • Little boy playing • Dog • Snow falling
14/6/15& 8& Method: Virtual Environment IWANN-2015 Results IWANN-2015 • They were no differences between classification errors in participants with first or second secuences 0 20 40 60 80 100 1 2 3 4 5 6 7 8 9 10 11 Percentage Participants Minimum Error Rate Phases A-B2-B1-C1-C2 Phases A-B1-B2-C2-C1
14/6/15& 9& Results IWANN-2015 • ANOVA with interface and distractors did not show significant differences, neither for interaction Interface x Distractors 0 20 40 60 80 100 B1 B2 C1 C2 Percentage Correct selection of commands Results IWANN-2015 • Control of navigation from participand did not depend on the type of interface they used • Control of navigation did not depend of distracting stimuli