Temporal Sensitivity Under Photopic and Scotopic Conditions Across the Cortical Visual Hierarchy

Abstract

Purpose: Behavioral and electrophysiological studies have shown that vision is slower under scotopic conditions (dark, activating only rods) than photopic conditions (light, activating only cones). However, slower scotopic processing cannot be solely explained by findings that rod signals are slower than cone signals, and it is unknown whether temporal processing differences persist in cortex. Flickering stimuli have previously been used in functional MRI (fMRI) studies to probe photopic cortical temporal sensitivity. This fMRI study investigates flicker sensitivity under photopic and scotopic conditions across the cortical visual hierarchy.

Methods: Fourteen participants viewed a stimulus flickering at six frequencies (2-10 Hz) under photopic and scotopic conditions during fMRI scanning. Retinotopic and high-level visual areas were delineated for each subject with population receptive field modeling (using a drifting bar) and a functional localizer (using images of objects).

Results: In most areas, higher mean activation was observed under photopic than under scotopic conditions. However, peak activation was higher only in V1 and ventral retinotopic areas. The pattern of change over frequencies differed between lighting conditions in retinotopic areas, but not in most high-level areas. Under scotopic conditions, the largest BOLD response was observed at low frequencies throughout visual cortex. Under photopic conditions, BOLD responses appeared largely unchanging across frequencies, with a trend towards preferring higher frequencies in V1.

Conclusions: Selectivity for lower frequencies under scotopic conditions was observed throughout visual cortex, in contrast to limited selectivity under photopic conditions. This low-frequency preference could allow more time for extracting information from sparse scotopic stimuli.

Figure 1.

Example demarcation of brain regions. Medial (left in each panel) and lateral (right in each panel) views of one subject's inflated right cortical surface (d = dorsal, v = ventral, a = anterior, p = posterior). (A) Colors indicate flicker-evoked activity (all frequencies > rest, FDR < 0.05): yellow—flicker-sensitive under photopic conditions; blue—flicker-sensitive under scotopic conditions; green—flicker-sensitive under both conditions. (B) Colored regions indicate ROIs defined using pRF or localizer experiment. GLMs were performed on the green vertices within each ROI.

Figure 2.

Beta weights across flicker frequencies (photopic—yellow; scotopic—blue). Results are shown separately for (A) V1, (B)–(F) extrastriate retinotopic areas, and (G)–(K) category-selective areas. Error bars indicate ± SEMs.

Figure 3.

Beta weight means, peaks, and slopes (photopic—yellow; scotopic—blue). (A) Beta weight means over frequencies, (B) beta weight peaks over frequencies, and (C) beta weight slopes over frequencies, shown for the different brain regions. Solid lines between bars represent a significant (P < 0.05) difference in a t-test between photopic and scotopic dependent variables. Asterisks represent a significant (P < 0.05) difference in a t-test of the dependent variable versus 0. Error bars indicate ± SEMs.