A novel ICA-based hardware system for reconfigurable and portable BCI

Abstract

Partially or completely paralyzed patients can benefit from Brain Computer Interface (BCI) in which a continuous recording of electroencephalogram (EEG) is required, operating some processing and classification to control a computer or other devices. Patients are so allowed to control external devices or to communicate simple messages through the computer, just concentrating their attention on codified movements or on a letter or icon on a digital keyboard. Using an on-purpose optimized spatial filtering technique in a BCI system, based on the ElectroEncephaloGraphic activity detection, enables to improve its accuracy, due to the explicit separation of the signal activity of interest from non-interesting signals. In this paper, a novel implementation of the spatial filtering ICA algorithm, onto a very performing and flexible hardware platform, is presented. The designed system enables the acquisition and the subsequent ICA elaboration of ElectroEncephaloGraphic signals in real-time, resulting so very effective for BCI applications. The efficiency of the implemented technique has been experimentally demonstrated.