VSS, May 13-18
Talk 1, 2:30 pm, 54.11
TALK 1 CANCELLED - A binocular synaptic network supports interocular response alignment in visual cortical neurons
Benjamin Scholl1, Clara Tepohl2, Connon Thomas2, Melissa Ryan3, Naomi Kamasawa2, David Fitzpatrick2; 1University of Pennsylvania, 2Max Planck Florida Institute, 3Baylor College of Medicine
In the visual system, a critical step in building coherent neural representations is combining signals from the two eyes. In carnivores and primates, binocular convergence first occurs in the primary visual cortex (V1) where individual cortical neurons respond selectively to sensory stimulation of one or both eyes. In all mammals investigated, most binocular V1 neurons exhibit matched preferences for stimuli viewed through each eye. While this alignment of response properties is a prerequisite for binocular vision, the synaptic basis is poorly understood. To understand the synaptic interactions underlying the binocular representation of stimulus orientation, we used two-photon in vivo imaging of calcium signals in dendritic spines on layer 2/3 binocular neurons in ferret visual cortex. Individual neurons with binocularly aligned somatic responses received a mixture of monocular and binocular synaptic inputs, which could potentially contribute to alignment. While contralateral monocular inputs were well-matched with somatic orientation preference, ipsilateral monocular inputs were not, indicating that monocular inputs alone cannot explain binocular somatic alignment. Within the population of binocular synaptic inputs, individual inputs exhibited varying degrees of interocular response alignment, but those with a high degree of tuning similarity (congruent) also had greater orientation selectivity and somatic specificity. Because they constitute only a third of a cell’s input population, we probed factors that could explain how congruent binocular inputs may impact somatic alignment. Using electron microscopy, we found that ultrastructural correlates of synaptic strength did not play a role. Instead, single-trial-based simulations of somatic aggregate input demonstrated that the relative fraction of active binocular congruent inputs predicts aggregate alignment to the somatic output. Altogether, our study suggests that coherent binocular responses of cortical neurons derive from biases in connectivity leading to functional amplification of aligned signals within a heterogeneous binocular recurrent network.
Acknowledgements: This work was supported by funding from NIH grants R01 EY011488 (D.F.), NIH grant K99 EY031137 (B.S.), the Max Planck Florida Institute for Neuroscience, and the Max Planck Society.
Talk 2, 2:45 pm, 54.12
Opponency versus normalization as the cause of interocular suppression in dichoptic masking
There are two models of the primary cause of interocular suppression (IOS) in binocular vision. The opponency model posits that IOS arises when imbalances in the two eyes’ signals are detected by opponency channels which then drive IOS. The normalization model posits that divisive normalization between the eyes’ signals, which occurs irrespective of whether the two eyes’ signals are imbalanced or not, drives IOS. We attempted to test between these two models in the context of a well-studied manifestation of IOS - dichoptic masking. A signature feature of dichoptic masking is that a mask in one eye increases the thresholds for detecting a test in the other eye more than it does a test in the same eye. Our masks and tests were horizontally-oriented, interocularly in-phase, 0.75 cpd gratings contained with a central 2 deg diameter test window. As expected we found a greater amount of opposite-eye compared to within-eye masking of the test stimulus. We then surrounded the mask-plus-test window with horizontally oriented, high contrast, 0.75 cpd, 6.5 deg diameter “surround masks” that were either interocularly in-phase or anti-phase. In a separate experiment we had showed that the anti-phase more than in-phase surround mask elevated thresholds for detecting an interocular difference in grating phase, in keeping with a role for the anti-phase surround mask as a desensitizer of opponency channels. We hypothesized that if IOS was driven by opponency channels the anti-phase more than in-phase surround masks would reduce the amount of signature dichoptic masking within the central mask/test window. However we found no evidence that the anti-phase surround masks reduced dichoptic masking more than the in-phase surround masks, thus providing no support for the opponency model of IOS and instead support for the normalization model.
Acknowledgements: Canadian Institute of Health Research grant #MOP 123349 given to F.K.
Talk 3, 3:00 pm, 54.13
Human stereovision is affected by adaptation in the monocular channels
Cherlyn Ng1 (), Martin Banks2, Randolph Blake3, Duje Tadin4, Geunyoung Yoon1; 1University of Houston, Texas, 2University of California (Berkeley), California, 3Vanderbilt University, Tennessee, 4University of Rochester, New York
Retinal images are the primary source of information for visual perception but subject to blur from the eyes’ imperfect optics. Improving the eyes’ optics should reduce blur and improve vision, but that was not always the case with stereovision. When tested with three-dimensional waves using adaptive optics, human participants with naturally good optics performed worse instead. The mean stereoacuity was 18.61arcsec through their own eyes but 33.34arcsec with the improved optics when tested with 1cpd waves. Likewise, stereoacuity was 28.22 and 53.82arcsec respectively, for 2cpd and 46.30 vs 85.50arcsec for 3cpd waves. This suggested that the visual system adapts to one’s own blur patterns to maximize stereovision. Evidence also came from replacing the participants’ own optics with someone else’s so that everyone now had the same blur patterns. Stereoacuity was also poorer in everyone compared to the person from whom the optics came (1cpd: 53.64arcsec [mean replacement] vs 8.46arcsec; 2cpd: 101.5 vs 12.91 arcsec; 3cpd: 182.34 vs 16.76arcsec). Did adaptation occur in monocular or binocular channels? We swapped the optics of the two eyes in comparison to eyes that received their own native blur. Swapping caused unfamiliarity in the monocular channels, but not in the binocular channels that still received the same combined blur and interocular difference. Stereoacuity was worse with the swap, indicating that adaptation had occurred in monocular channels (1cpd: 28.4arcsec with the native optics vs 43.4arcsec swapped; 2cpd: 53.6 vs 123.0arcsec; 3cpd: 144.5 vs 221.9arcsec). The detriment grew with interocular difference (1cpd: r=0.91, p=0.01; 2cpd: r=0.97, p=<0.01; 3cpd: r=0.91, p<0.01) as expected because swapping would matter only if the blur patterns in the two eyes were different. These results indicate a component of adaptation that happens before binocular combination towards the blur patterns of individual eyes.
Acknowledgements: This work was supported by NIH grants EY014999 and P30 EY001319, and funding from Research to Prevent Blindness (RPB).
Talk 4, 3:15 pm, 54.14
Neurophysiology and psychophysical studies of the stereo contrast paradox
Typically, stereoscopic disparity thresholds decrease for lower stimulus contrast. However, for low frequency luminance gratings, raising the contrast in just one eye impairs disparity thresholds more than if contrast is lowered in both eyes. This phenomenon, named stereo contrast paradox, is absent for high spatial frequency gratings and for stimuli that are frequency and orientation broadband. Using random dot (RDS) and random line stereograms (RLS), that are broadband in frequency and respectively broadband and narrowband in orientation, we investigated the activity of disparity-selective neurons in macaques primary visual cortex (V1) that perform the initial computations for stereopsis, and combined neuronal recordings with psychophysical measurements in primates and humans. Neurally, for RDS, disparity modulation is reduced more severely when contrast is lowered in both eyes compared to when it is lowered in just one, consistent with the absence of the paradoxical effect (median reduction factor 0.2 vs 0.4, pVal <0.001 sign rank test). However, for RLS, firing rates and disparity sensitivity were higher when contrast was low in both eyes than when it was mismatched between them, consistent with the presence of the paradox (0.7 vs 0.6, pVal < 0.001). To compare our physiology and psychophysics results, we introduced a neurometric discriminability d’ metric. Nevertheless, psychophysically, we found no paradox for either RDS or RLS in either humans or macaques, and replicated previous psychophysics results for RDS and gratings at different spatial frequencies. The discrepancy suggests that stereoacuity thresholds cannot be explained by a fixed linear decoder applied to V1.
Acknowledgements: Wellcome Trust, UK and NIH, NEI, USA
Talk 5, 3:30 pm, 54.15
Primate monocular vision is intrinsically unstable: a side-effect of binocular homeostasis
Alexandre Reynaud1 (), Kévin Blaize2, Fabrice Arcizet3, Pierre Pouget4, Serge Picaud3, Frédéric Chavane2, Robert Hess1; 1McGill University, 2Institut de Neurosciences de la Timone (INT) - CNRS, Aix-Marseille Université, 3INSERM, CNRS, Institut de la Vision, Sorbonne Université, 4INSERM, CNRS, Institut du Cerveau et de la Moelle épinière, Sorbonne Université
Short-term deprivation of one eye of 1-2 hours in human adults results in reciprocal changes in the interocular balance that last 30-90 minutes, assumed to be the consequence of a change in the contrast gain within the binocular circuitry. What is not known is how these changes emerge and with what dynamic. One possibility is that they are initiated once the patch is removed. The alternative is that they result from a dynamic build-up from the moment the patch is applied. In this study, we measure the monocular dynamics of the non-deprived eye activation during the deprivation of the other eye in humans, using psychophysics, and in awake non-human primates, using functional ultrasound imaging (fUS) of the primary visual cortex (V1). In humans, we show that the contrast threshold increases overtime during the deprivation of the other eye. This observation indicates that the patching effect slowly builds up over the deprivation period at a binocular site. In the awake macaque, we show that during monocular deprivation the cerebral blood volume (CBV) increases in the non-deprived eye ocular-dominance bands, but not the deprived eye bands with similar dynamics as observed in human psychophysics. We show that human behavioral results can be predicted by a simple interocular normalization model that uses non-human primate CBV dynamics as a descriptor of the normalization pool activity. In conclusion, our results highlight a hitherto unknown feature of primate vision: monocular vision per se is intrinsically unstable.
Acknowledgements: Fondation de France grant to KB, CIHR grant #228103 to RFH
Talk 6, 3:45 pm, 54.16
Discomfort associated with the (un)natural statistics of VR gaming headsets
Avi M. Aizenman1 (), George A. Koulieris2, Agostino Gibaldi1, Vibhor Sehgal1, Dennis M. Levi1, Martin S. Banks1; 1Herbert Wertheim School of Optometry & Vision Science at the University of California, Berkeley, 2Department of Computer Science at Durham University
Human binocular vision is adapted to statistical regularities in the natural environment such that depth perception and binocular eye movements are precise, fast, and comfortable. We measured for the first time the statistics of fixation, disparity, and vergence-accommodation conflict in virtual reality (VR) and compared them to statistics when viewing the natural world. Specifically, we collected eye-position and 3d-scene-geometry data during tasks performed in VR (first-person shooter games, environmental simulation, and beat/rhythm games) and the natural environment (making a sandwich, ordering coffee, taking indoor and outdoor walks, and editing text). We found that gaze is biased towards straight ahead and farther distances in VR relative to the natural environment (VR: 1.25m; natural environment: 0.67m). We also measured the statistics of the vergence-accommodation conflict (VAC) and found that ~80% of fixations in VR produce significant conflict. From this, we determined the optimal screen distance to minimize discomfort due to VAC. In the natural environment, the vertical horopter and natural-disparity statistics exhibit a top-back pitch, a pattern not present in the VR environment. Specifically, the VR produces significantly smaller near (crossed) disparities in the lower visual field than the natural environment, a difference as large as 900arcsec. Finally, we tested whether observers prefer VR content that is consistent as opposed to inconsistent with the statistics of the natural world. We found that content that violates the top-back pitch of the natural world generates more discomfort and reduced performance. We conclude that the mismatch between the statistics of the VR and natural environments leads to discomfort and reduced performance in VR headsets. Our findings inform improvements to VR headset design and content in order to be more consistent with the statistics of the natural environment.
Acknowledgements: The Center for Innovation in Vision and Optics (CIVO) at UC Berkeley for funding this project, and HTC for loaning us headsets for the experiment.
Talk 7, 4:00 pm, 54.17
Binocular viewing geometry shapes neural processing of slanted planes: Results from theoretical V1 modeling and human psychophysics
An important component of real-world scene perception is the estimation of 3D surface orientation (slant and tilt). However, neither the neural circuits that compute 3D orientation nor the supporting computations have been fully uncovered. To help address that gap, we apply a projective geometry framework to the study of both the neural processing and perception of 3D orientation, building on recent work that suggests the mapping of the 3D environment onto the two retinae shapes both the tuning of neurons in area MT to motion-in-depth and corresponding patterns of perceptual errors in human psychophysics (Bonnen et al., 2020). The projection of slanted planes results in interocular disparities in retinal orientation, and interestingly, binocular V1 neurons can have different monocular orientation preferences (Bridge & Cumming, 2001), which could make them sensitive to orientation disparities. Therefore, we have constructed an encoding model that predicts responses of binocular V1 neurons to the orientation disparities produced by a stimulus plane with varying slant, given each theoretical neuron’s monocular orientation preferences and ocular dominance level. The resulting slant tuning curves are noncanonical in shape, do not tile slant evenly, and are strongly affected by viewing distance. Notably, at greater viewing distances, extreme slants (beyond approximately +/- 70 degrees) are encoded by steep portions of the curves, but moderate and low slants are encoded by relatively flat portions. As expected based on those observations, our model population decoder performs best at extreme slants, and performance degrades as slant approaches zero and as viewing distance increases. We additionally observed these patterns in human perceptual performance on a slant discrimination task. These results support the theory that the neural processing of 3D scene components is shaped by projective geometry and represent a step toward advancing understanding of the computations supporting 3D orientation perception.
Acknowledgements: NIH NEI 5R01EY020592 (LKC, ACH); NSF GRFP (supporting SMS)