Temporal Patterns Analysis in EEG Data using Sonification

Abstract

Brain computer interfacing is a direct communication pathway between human and computer. The area is innovative and under active research as a subset of human computer interaction. Brain computer interfacing typically utilize electroencephalogram (EEG) data to apply machine learning and or computational intelligence techniques to trigger actions. Electroencephalogram data is a multivariate time series data whose visualization suffers from problems like cluster overlapping and high dimensionality. In this paper we present a natural alternative to visualization i.e. sonification. Thus combining the two novel area of research to explore and get better results over the conventional approaches. Sonification is the use of non-speech audio to convey information. We have presented a technique to sonify alpha, beta and gamma bands present in the Electroencephalogram data. Experiments have been performed on EEG data of a controlled subject and the results have been discussed at the end.