Learning in closed-Loop brain–machine interfaces: modeling, experimental validation, and applications
Rodolphe Héliot

November 3th, 2010 at 10:30 am
Seminar room - LIRMM

Abstract
A brain–machine interface (BMI) is a direct communication pathway between the brain and an external artificial actuator, such as a computer or a robot. Closed-loop operation of a brain–machine interface relies on the subject’s ability to learn an inverse transformation of the plant to be controlled. We propose a model of the learning process that undergoes closed-loop BMI operation and explore its properties to show that it is able to learn an inverse model of the controlled plant. Comparisons between model predictions to actual experimental neural and behavioral data from nonhuman primates demonstrate high accordance of the model with the experimental data. Applying tools from control theory to this learning model can help in the design of optimal neural information decoders which will maximize learning speed for BMI users. An application to neuron-loss compensation will be presented, showing dramatic performance improvements.