The Alignment of Systemic Low Frequency Oscillations with V1 Retinotopic Organization

Abstract

It is now well established that the BOLD signal contains systemic low frequency fluctuations (sLFOs, below 0.2 Hz) that are due to global blood-borne signals propagating through the vasculature (Tong et al., 2013). These sLFO signals are particularly noticeable in primary sensorimotor, auditory and visual networks due to their high blood capillary density and/or vascular density (Harrison et al., 2002). Our goal was to examine the effect of these sLFOs on estimated population receptive field (pRF) maps within early visual areas. We used standard techniques (Dumoulin and Wandell, 2008) to estimate pRFs (the Gaussian in visual space that best predicts each voxel’s time course) for 8 participants from the Human Connectome Project 7T Retinotopy dataset (using minimally processed 2mm-32k retinotopy data; the stimuli were moving bars, rotating wedges, contracting/expanding rings) both before and after removing estimated sLFOs. SLFOs were estimated by selecting highly vascularized regions as initial seeds for a recursive procedure that tracks the evolution of sFLO-driven time-lagged correlations through the brain (rapidtide, Frederick 2016). We found that shared time-course variances due to sLFOs were stronger across regions of the cortex that shared the same eccentricity. This may be the result of sLFO’s following the vasculature that symmetrically stems from the posterior cerebral artery (Tong and Frederick, 2014), and thereby aligns with eccentric organization. Thus, vascular architecture may explain previous results showing strong correlations between voxels that share the same eccentricity (Butt, Benson, Datta, & Aguirre, 2015; Bao and Tjan, 2009; Arcaro et al. 2015). Fortunately, because local correlations are preserved, removing sLFOs does not change the position of estimated pRFs.

Acknowledgement: Washington Research Foundation National Eye Institute and Office of Director, Office of Behavioral and Social Sciences Research