Page 162 - Handbook of Biomechatronics

Biomechatronic Applications of Brain-Computer Interfaces 159

Fig. 6 A person uses a 7-degree-of-freedom rehabilitation robot while a brain-computer
interface monitors their mental workload. DF ¼ degrees of freedom: DFs 1–3 are in the
shoulder (partially obscured by user), DF 4 is in the elbow, DFs 5 and 6 are in the lower
arm (lower arm pronation/supination and wrist flexion/extension), and DF 7 is the hand
opening/closing module; EEG ¼electroencephalogram. (From the author’s joint research
with Prof. Jos e del R. Millán and Dr. Tom Carlson, Ecole Polytechnique Federale de Lau-
sanne, Switzerland.)

success-rate-based system, it is unclear whether this improvement is large
enough to justify the additional complexity and unobtrusiveness. Similar
EEG-based prototypes have been developed for, for example, computerized
reading tutors (Chang et al., 2013) and serious games that teach fire safety
(Ghergulescu and Muntean, 2014), but have also not yet shown clear
benefits.
Difficulty adaptation is not limited only to education and training. It can
also be used in computer games simply for entertainment: making the game
more fun by ensuring that the player is neither bored nor frustrated. An
important study in this area was conducted by Chanel et al., who found that
player engagement in a game of Tetris can be estimated from EEG with a
reasonable accuracy; furthermore, they showed that EEG is a better indica-
tor of engagement than autonomic nervous system responses (Chanel et al.,
2011). Ewing et al. (2016) later built on this knowledge to design a BCI that
estimated player engagement during Tetris based on frontal and parietal