Spatial Distribution of Attentional Modulation at Columnar Resolution in Macaque Area V4

Abstract

Attention to a location in a visual scene affects neuronal responses in visual cortical areas in a retinotopically specific manner. Optical imaging studies have revealed that cortical responses consist of two components of different sizes: the stimulus-nonspecific global signal and the stimulus-specific mapping signal (domain activity). It remains unclear whether either or both of these components are modulated by spatial attention. In this study, to determine the spatial distribution of attentional modulation at columnar resolution, we performed cerebral blood volume (CBV)-based optical imaging in area V4 of monkeys performing a color change detection task in which spatial attention was manipulated. We found that spatial attention enhanced global signals of the hemodynamic responses, but did not affect stimulus-selective domain activities. These results indicate the involvement of global signals in neural processing of spatial attention. We propose that global signals reflect the neural substrate of the normalization pool in normalization models of attention.

Diversity of Feature Selectivity in Macaque Visual Cortex Arising from a Limited Number of Broadly Tuned Input Channels

“…Our study also establishes the value of observing the OI signal at different spatial scales to reveal underlying processes as has been pointed out in an fMRI study done on the primary visual cortex of cats and humans (Swisher et al, 2010) and as recently demonstrated also for macaque area V4 (Tanigawa et al, 2016). Furthermore, our findings resolve an apparent discrepancy in the earlier literature about orientation anisotropies in the visual system.…”

Representation of Cone-Opponent Color Space in Macaque Early Visual Cortices

“…Detailed imaging and data analysis methods have been described previously ( Tanigawa et al, 2010 , 2016 ; Hu et al, 2020 ). Prior to the imaging, the chamber cap was opened under sterile conditions, and the exposed cortex was stabilized with warm 3% agarose and a glass disk.…”

“…For each stimulus presentation, the average of frames obtained during the pre-stimulus period ( R0 , 1.0–0 s before the stimulus onset) and the average of frames during the stimulus period ( R1 , 1.0–3.5 s after the stimulus onset) were calculated on a pixel-by-pixel basis, and then a map of reflectance change ΔR/R (response map) was generated as (R1-R0)/R0 . To extract locally evoked reflectance changes (mapping signals: ∼0.5 mm) from large-scale non-stimulus-specific changes (global signals: several millimeters or more; Frostig et al, 1990 ; Tanigawa et al, 2016 ), each response map was convolved with a 1.0 mm × 1.0 mm median filter and subtracted from the original map (high-pass filtering). Using response maps, single-condition maps were generated by subtracting the blank condition from each stimulus condition, and subtraction maps were generated by calculating the difference between two stimulus conditions ( Tanigawa et al, 2010 ).…”