277 | Meaningful feedback: user-centered coadaptive brain-computer interfaces

Theoretical and Computational Neuroscience

Author: Denise Gabriela Nigro | Email: lic.denisenigro@gmail.com


Denise Gabriela Nigro , Catalina Galván , Paula Saavedra , Victoria Peterson

1° Denise Gabriela Nigro
2° Catalina Galván

Brain Computer Interfaces (BCIs) are systems that measure the brain activity and convert it into artificial control outputs. For motor neurorehabilitation, these technologies can be used to transform movement attempt/imagination into visual and/or somatosensory feedback. Experimental evidence shows that up to 50% of people are not able to command a BCI in the first session. Of these, about 20% are unable to gain control of the BCI throughout session of usage. In this work we design a user-centered stimulation protocol which provides feedback based on operant conditioning, taking into account optimal levels of attention and motivation.
The protocol is designed as a videogame, considering contributions from cognitive-behavioral psychology. The Neurofeedback is based on the outputs of an algorithmic solution that can adapt to the user’s brain changes. The user needs to focus on the imagination of the grasping movement to reach a bag full of money. The feedback is designed so as to be adjusted as the user’s performance improves, aiming to achieve the greatest possible autonomy from the algorithm. The reinforcers will gradually decrease depending on the subject’s performance. The level of fatigue will be monitored to stop or/and pause the game when necessary. A diary of progress will also be delivered to the user by the end of each BCI session. We expect that by means of this percentage of inefficiency in BCI users disminish.