Population receptive fields in human visual cortex measured with subdural electrodes

Abstract

PURPOSE. Electrophysiological methods in animal models have been used to identify receptive field properties of neurons within retinotopic maps. More recently functional magnetic resonance imaging (fMRI) methods in human have been used to estimate population receptive fields (pRF) (Dumoulin and Wandell, 2008; Kay et al., 2008). Following Yoshor et al. (2007), we developed an efficient method of electrocorticography (ECoG) using subdural electrodes in pre-surgical clinical subjects to estimate pRFs in human visual cortex. These measures bridge human fMRI and animal electrophysiological studies. METHODS. Two patients with implanted intracranial electrodes (2-mm surface diameter) viewed a flickering contrast pattern through a bar aperture that swept across the visual field 8 times (4 cardinal, 4 diagonal directions; 96 seconds total). The contrast pattern flickered at 7.5 Hz, creating a steady-state ECoG response with power concentrated at twice the stimulus frequency (15 Hz). For each electrode a time-series of the time-varying 15-Hz amplitude was extracted, and modeled using an isotropic 2D-Gaussian pRF. RESULTS. The pRF model fit occipital electrodes' time-series well, explaining up to 83% of the variance. The pRF parameters were similar to those obtained from fMRI; for example, the pRF size increased with eccentricity and was larger in extrastriate regions than in electrodes near the occipital pole. The signal response latencies were estimated from the phase of the response to full-field flicker (separate runs). We observed robust position-dependent latency effects, ranging from 10–40 ms delay relative to responses near the occipital pole. CONCLUSION. Population receptive fields can be estimated using ECoG. There is good agreement between fMRI and ECoG measures despite significant differences in their physiological bases. The ECoG data provide latency information that is unavailable in the fMRI responses. This is a valuable method for probing population-level spatiotemporal properties of receptive fields in human visual cortex.