Decoding Steady State Visually Evoked Potentials for BCI applications using canonical correlation analysis and multivariate synchronization index.

Brain computer interfaces is nascent field of research that at present is focussed on helping people with neurological disorders or neurodegenerative dieseases that inhibit movement. Decoding steady state visually evoked potentials has shown considerable promise to help build this technology. These signals have been used for spellers, wheelchair control and neural engineering. In this project we have implemented two algorithms for decoding these signals. We have further investigated the utility of signals taken from a single channel (Oz in the 10-20 system) from the visual cortex.

Since EEG signals are inherently noisy in nature, we have included overlapping windows in our analysis. We have implemented two decoding algorithms called Canonical Correlation Analysis(CCA) and Multivariate Synchronization index. This repository contains the implementation for multivariate synchronization index. The results of this research has been published in the IET signals processing journal in a paper titled: "Investigation of multiple frequency recognition from single channel steady state visually evoked potential for efficient brain computer interface application"