class="head no_bottom_margin" id="sec1title">IntroductionA remarkable property of the visual system is its feat to provide us with stable vision despite continuously changing retinal input induced by our movements of eyes, head, and body. This feat appears especially intriguing as the majority of visual areas are organized retinotopically, yet stability requires integration of visual input with cues from other modalities. Although the integration of eye movements with retinal signal has been studied extensively in both monkeys (, , , , , ) and humans (, , , ), the integration of visual signal with voluntary head movements remains barely studied at the level of neocortex (see , for subcortical function).In macaques and humans, previous studies examining cortical function focused on passive head motion or artificial vestibular stimulation to examine visual-vestibular integration (, , , href="#bib24" rid="bib24" class=" bibr popnode">Frank et al., 2014, href="#bib23" rid="bib23" class=" bibr popnode">Frank et al., 2016, href="#bib2" rid="bib2" class=" bibr popnode">Billington and Smith, 2015). However, active gaze shifts beyond eye movements also involve head rotation (href="#bib38" rid="bib38" class=" bibr popnode">Land, 1992). In fact, gaze change commands reach eye and head effector muscles at the same time (href="#bib3" rid="bib3" class=" bibr popnode">Bizzi et al., 1971), and human observers compensate for eye- and head-induced self-motion with equal precision (href="#bib14" rid="bib14" class=" bibr popnode">Crowell et al., 1998). Notably, however, despite this prominent role for head motion in visual stability almost nothing is known about which visual processing stages integrate head motion signals with retinotopic representations as technical limitations have hindered human neuroimaging to study the neural underpinnings of voluntary head movements.We recently circumvented these limitations and introduced an approach that allows participants to move their heads during fMRI scanning by exploiting the delay of several seconds between neural processing and blood-oxygen-level-dependent (BOLD) signal (see href="/pmc/articles/PMC6153141/figure/fig1/" target="figure" class="fig-table-link figpopup" rid-figpopup="fig1" rid-ob="ob-fig1" co-legend-rid="lgnd_fig1">Figures 1A–1C) (href="#bib46" rid="bib46" class=" bibr popnode">Schindler and Bartels, 2018). We constructed a custom-built air pressure-based head stabilization system that permitted head rotation during trials, but stabilized head position during data acquisition. Observers wore head-mounted magnetic resonance-compatible goggles while head movement was tracked online. This allowed generation of visual stimuli that could be modulated by head motion in real time (href="#bib46" rid="bib46" class=" bibr popnode">Schindler and Bartels, 2018). In two conditions, observers viewed approaching visual flow that was modulated by head motion. A congruent condition simulated a scenario of constant forward motion where head rotation resulted in looking around while being driven along a straight road. In the incongruent condition, observers performed identical head rotations, but visual consequences of head rotation were inversed such that visual and extra-retinal cues did not combine in any meaningful way but retinal motion was matched to the congruent condition. In both conditions, a demanding letter detection task assured fixation. Based on this paradigm we previously examined the integration of head movements and visual signals in a network of areas with established vestibular input. Particularly, a contrast between congruent and incongruent conditions revealed evidence consistent with the multi-modal integration of visual cues with head motion into a coherent “stable world” percept in the parietal operculum and in the anterior part of the parieto-insular cortex. This also applied for a subset of visual motion-responsive areas such as human medial superior temporal area (MST) (at uncorrected level), the dorsal part of the ventral intraparietal area (VIP), the cingulate sulcus visual area (CSv), and a region in the precuneus (Pc) (href="#bib46" rid="bib46" class=" bibr popnode">Schindler and Bartels, 2018). However, the important question whether retinotopic cortex and especially areas V3A and V6 play a role in visual stability during voluntary head movement remained open.href="/pmc/articles/PMC6153141/figure/fig1/" target="figure" rid-figpopup="fig1" rid-ob="ob-fig1">class="inline_block ts_canvas" href="/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=6153141_gr1.jpg" target="tileshopwindow">target="object" href="/pmc/articles/PMC6153141/figure/fig1/?report=objectonly">Open in a separate windowclass="figpopup" href="/pmc/articles/PMC6153141/figure/fig1/" target="figure" rid-figpopup="fig1" rid-ob="ob-fig1">Figure 1Illustration of Visual Stimuli and Head-Rotation Task, BOLD Signal Acquisition during a Trial, and Experimental Paradigm(A) Observers performed voluntary head rotations while being approached by a simulated 3D dot cloud in both congruent and incongruent conditions. Head rotations in the congruent condition lead to cloud rotation in opposite direction (-α) to the observer's head (α), as would be experienced when moving forward in a stable environment and looking around. In the incongruent condition, the cloud and head rotated in the same direction (α), resulting in perceptually arbitrary motion of the environment. Note that retinal flow as well as head motion were matched in both conditions.(B) Model of the evoked BOLD time course as predicted by the paradigm in (C). Stimulus presentation and active head movements induced BOLD signals during the trial phase (green shade) while the slow dynamics of the BOLD signal allowed acquisition of these responses even after stimulus offset, at a time when the observer's head was stabilized (acquisition phase, red shade).(C) Each trial started with an instruction phase when air cushions were emptied. In the trial phase, green arrowheads guided the observer's head rotation. In the acquisition phase, air cushions were inflated again to record BOLD responses. Observers performed a demanding fixation task across all phases and conditions, except the instruction phase (see href="#sec4" rid="sec4" class=" sec">Methods).
展开▼
