Summer Research Fellowship Programme of India's Science Academies

Numerical tools to infer functional Brain connectivity from EEG data

Sreelakshmi Rajendrakumar

School of Biotechnology, National Institute of Technology Calicut, Calicut 673601

Dr. Shivakumar Jolad

Assistant Professor, Department of Physics, Indian Institute of Technology Gandhinagar, Gandhinagar 382355


In this work numerical tools are used for frequency domain analysis of time series data obtained using EEG. The time series used in this study is intended to give an insight into the difference in connectivity patterns of epileptic brain and normal brain. The time series are subjected to Fourier transformation for obtaining the frequency components that make them up which is in turn subjected to the computation of power spectral densities of every time series and cross spectral densities between different time series. The coherence matrix that is necessary to carry out the analysis is obtained using power spectral density and cross spectral density value of the time series. This coherence study can be extended to extract functional brain connectivity patterns that can help in differentiating between brain states, whether normal or diseased.

Keywords: EEG, time series data, Fourier transform, spectral density, coherence

Written, reviewed, revised, proofed and published with