The Stockwell Transform in Studying the Dynamics of Brain Functions

The dynamics of brain functional activities make time-frequency analysis a powerful tool in revealing its neuronal mechanisms. In this paper, we extend the definition of several widely used measures in spectral analysis, including the power spectral density function, coherence function and phaselocking value, from the classic Fourier domain to the time-frequency plane using the Stockwell transform. The comparisons between the Stockwell-based measures and the Morlet wavelet-based measures are addressed from both theoretical and numerical perspectives. The Stockwell approach has advantages over the Morlet wavelet approach in terms of easy interpretation and fast computation. A magnetoencephalography study using the Stockwell analysis reveals interesting temporal interaction between contralateral and ipsilateral motor cortices under the multi-source interference task.