Orientation Anisotropies in Human Visual Cortex

Abstract

senting the orientation of features in the visual image is a fundamental operation of the early cortical visual system. The nature of such representations can be informed by considering anisotropic distributions of response across the range of orientations. Here we used functional MRI to study modulations in the cortical activity elicited by observation of a sinusoidal grating that varied in orientation. We report a significant anisotropy in the measured blood-oxygen leveldependent activity within visual areas V1, V2, V3, and V3A/B in which horizontal orientations evoked a reduced response. These visual areas and hV4 showed a further anisotropy in which increased responses were observed for orientations that were radial to the point of fixation. We speculate that the anisotropies in cortical activity may be related to anisotropies in the prevalence and behavioral relevance of orientations in typical natural environments. Although such studies considered pattern orientation independent of position in the visual field, anisotropic response distributions have been reported across the early human visual system when pattern orientation is specified relative to the local visual field meridian (Fig. 1). Specifically, responses are enhanced when the local pattern orientation is coincident with the angular meridian (radial) compared with when it is tangential (Clifford et al. 2009; Sasaki et al. 2006). An important consequence of this meridian-relative anisotropy is that it confers difficulties in the interpretation of putative anisotropies in field-independent pattern orientation, particularly when the stimulus (Furmanski and Engel 2000; Furmanski et al. 2004) or analysis (Serences et al. 2009; Yacoub et al. 2008) is restricted to a subregion of the visual field. In this circumstance, inferences about anisotropies in field-independent orientation can be confounded with those in meridian-relative orientation; for example, a stimulus restricted to the vertical visual field meridian will produce responses in which only the vertical fieldindependent orientation is also a radial meridian-relative orientation.The aim of this study was to measure the distribution of visual system responses to varied pattern orientation by considering both field-independent and meridian-relative indices. We used fMRI to measure the blood-oxygen level-dependent (BOLD) response to a sinusoidal grating that varied in orientation and to map the preferred visual field meridia from within the early retinotopic regions of human visual cortex. We report the presence of significant anisotropies to both field-independent and meridian-relative orientation from within visual areas V1, V2, V3, and V3A/B. Although the form of meridianrelative anisotropy is consistent with previous reports of a preference for radial orientations, the anisotropy in fieldindependent orientation is characterized by a reduced response to horizontal orientations. M E T H O D S SubjectsFour experienced psychophysical observers participated in this study. Each subject received a re...

Gain control in the response of human visual cortex to plaids

“…The orientation of the plaid is defined as that midway between the two grating components. The cortical responses to gratings resemble the results of Mannion et al (2010); in V1, the response to the gratings is greatest at obliques, weakest at horizontal, and intermediate for vertical orientations. Furthermore, in agreement with experiment 1, the plaid stimuli evoke a greater BOLD response than the gratings from all cortical areas.…”

“…However, there are a number of plausible models described below that cannot be distinguished by these data. To address this issue, we conducted a second experiment comparing orientation response profiles for plaids with those for their component gratings.We have recently demonstrated that several of the areas of human early visual cortex have anisotropic orientation tuning (Mannion et al 2010; see also Saproo and Serences 2010;Swisher et al 2010). The BOLD response in V1 is typically greatest at the obliques, weaker at vertical orientations, and weakest at horizontal orientations.…”

“…To equate apparent motion signals within the grating and plaid conditions, the order of the grating stimuli was constrained such that the change in orientation from 1 stimulus to the next was never Ͼ45°. Grating and plaid blocks were matched in the distribution of orientations to avoid potential biases introduced by anisotropies in response as a function of spatial orientation (Mannion et al 2010;Sasaki et al 2006).Attention and eye movements were controlled by requiring observers to perform a demanding dimming task at fixation; throughout all blocks, participants were required to indicate with a button press whenever there was a brief luminance decrement of the fixation marker presented at the center of the display.Second experiment. Stimuli were sinusoidal gratings with a spatial frequency of 3.35 cycles/° (Fig.…”

Norms are Not the Norm: Testing Theories of Sensory Encoding Using Visual Aftereffects

“…This would be analogous to the anisotropies found in orientation-tuned channels, that are thought to underlie our superior discrimination ability about vertical and horizontal relative to oblique orientations (Girshick et al, 2011;Li et al, 2003;Mannion, McDonald, & Clifford, 2010;Storrs & Arnold, 2015a). Badcock et al (2014) recently communicated via a conference abstract that they had measured spatially-contingent aspect ratio aftereffects, and also concluded that aspect ratio adaptation involves a local repulsion.…”

Psychophysical analysis of three separate anisotropic patterns in vision : putting visual reference frames in conflict to study the class 1 and 2 oblique effects and the horizontal effect.

“…M. used functional near-infrared spectroscopy to measure the oxygenated hemoglobin response and also found the magnitude of response to be greatest in response to obliquely oriented gratings, though the differences were only significant in the left occipital lobe. This anisotropic response was also found in a greater blood-oxygen-level-dependent (BOLD) signal for obliques by Mannion, McDonald and Clifford (2010) in V1, V2, V3 and V3A/B using oriented sinusoidal gratings as stimuli. Nasr and Tootell (2012) tested the idea that the orientation bias arises at a higher visual area than V1/V2.…”