Does spike synchrony provide a better code of stimulus angle than average firing rate?

Abstract

We compared the degree of orientation discrimination coded by neuronal firing rate with that coded using spike synchrony. A 100 electrode array was used to record single cell activity in primary visual cortex (V1) of two prosimian bush babies (Otolemur garnetti). Electrodes were verified to be in layers 2–4 of V1 based on reconstructions from cytochrome oxidase stained sections. Cell responses were characterized using 18 orientations of drifting sinewave grating stimuli presented at 60% contrast at the preferred spatial (0.5c/deg) and temporal (2 Hz) frequency. 306 cell pairs with overlapping receptive fields were examined for synchrony using the joint peristimulus time histogram (JPSTH) method (Aertsen 1989). As described earlier in cat area 17 (Samonds 2003), we found that neuronal synchrony conferred a distinct advantage over firing rate for discrimination between small differences in orientation (10°) but not larger differences. At 10° synchrony drops 4.16%/degree and combined firing rate drops 2.28%/degree. After normalization, comparison shows that in this range synchrony offers an advantage in orientation discriminability of 67±9% over firing rate for the same cell pairs. In the 10–60° range, synchrony drops 1.21%/degree and combined firing rate drops 2.89%/degree. For these larger differences, firing rate offers 139±15% more discriminability than synchrony. These initial results suggest that synchrony is important for fine orientation discrimination in primate V1 but not for discriminating between larger orientation differences. These results strengthen the argument that neural synchrony is important to the coding of stimulus detail. (Supported by EY01778, EY08126, HD15052(VAC), EY014680-03(ABB)).