Time-course reconstruction of neural activity for multiples simultaneous sources

Abstract

Abstract-In this paper a new methodology for time-course reconstruction of neural activity is proposed. Proposed approach allows finding the locations and waveforms (time-course activity reconstruction) of inner brain sources, that could be used for source-based noninvasive brain computer interface technologies. Proposed framework for source-reconstruction considers a dynamic state space model which allows successfully the reconstruction of time-course activity at each source. A comparative analysis is performed against dynamic Thikonov methods, where the proposed approach obtain the minimum reconstruction error.