A statistical comparison of the spatial distribution of ASSR sources across age

Changes in auditory temporal processing may affect speech perception in ageing persons. Temporal processing is mediated by stimuli-induced synchronized activity in the central auditory system. The auditory steady-state response (ASSR) is a non-invasive measure of synchronized neural activity, induced by amplitude modulated (AM) acoustic stimuli. Different brain areas are involved in the generation of ASSRs. In this study we introduce a statistical framework to investigate possible effects of age on the spatial distribution of ASSR sources (generators) in the brain. ASSRs were obtained of young and older persons after presenting AM noises at 4 Hz and 40 Hz, two modulation frequencies that are present in continuous speech. Independent component analysis (ICA) was used as a linear decomposition to separate the brain source signals mixed by volume conduction in electroencephalography (EEG) recordings. An equivalent current dipole was then fitted for each source using a 4-shell spherical head model. The sources with significant responses to the presented modulation frequencies were determined based on Hotelling's T-squared test. The spatial distribution of these components was modeled by a Gaussian mixture model (GMM). Finally, the Kullback Leibler divergence was applied to provide a comparison between the GMM of ASSR sources in young and older persons. The result of the comparison showed a larger difference between young and older persons for 4 Hz ASSR sources compared to 40 Hz ASSR sources. This suggests that advancing age might have a larger effect on the spatial distribution of the 4 Hz ASSR sources compared to the 40 Hz ASSR sources.status: publishe