Shape perception via a high-channel-count neuroprosthesis in monkey visual cortex

Blindness affects 40 million people across the world. A neuroprosthesis could one day restore functional vision in the blind. We implanted a 1024-channel prosthesis in areas V1 and V4 of the visual cortex of monkeys and used electrical stimulation to elicit percepts of dots of light (called phosphenes) on hundreds of electrodes, the locations of which matched the receptive fields of the stimulated neurons. Activity in area V4 predicted phosphene percepts that were elicited in V1. We simultaneously stimulated multiple electrodes to impose visible patterns composed of a number of phosphenes. The monkeys immediately recognized them as simple shapes, motions, or letters. These results demonstrate the potential of electrical stimulation to restore functional, life-enhancing vision in the blind.

Protein expression of MEF2C during the critical period for visual development in vervet monkeys

During the early development of the visual cortex, there is a critical period when neuronal connections are highly sensitive to changes in visual input. Deprivation of visual stimuli during the critical period elicits robust anatomical and physiological rearrangements in the monkey visual cortex and serves as an excellent model for activity-dependent neuroplasticity. DNA microarray experiments were previously performed in our lab to analyze gene expression patterns in area V1 of vervet monkeys subjected to monocular deprivation (MD). An interesting candidate identified in its screen was myocyte enhancer-binding factor 2C (MEF2C), a transcription factor linked to neuronal survival. Consistent with the microarray data, we show that there is a qualitative increase in MEF2C protein expression in area V1 of infant as compared to adult vervet monkeys. Our results suggest that the regulation of neuronal survival is one of the molecular mechanisms underlying the critical period for visual cortical neuroplasticity.

Curvature domains in V4 of Macaque Monkey

An important aspect of visual object recognition is the ability to perceive object shape. How the brain encodes fundamental aspects of shape information remains poorly understood. Models of object shape representation describe a multi-stage process that includes encoding of contour orientation and curvature. While modules encoding contour orientation are well established (orientation domains in V1 and V2 visual cortical areas), whether there are modules for curvature is unknown. In this study, we identify a module for curvature representation in area V4 of monkey visual cortex and illustrate a systematic representation of low to high curvature and of curvature orientation, indicative of curvature hypercolumns in V4. We suggest that identifying systematic modular organizations at each stage of the visual cortical hierarchy signifies the key computations performed.SignificanceWe use intrinsic signal optical imaging in area V4 of anesthetized macaque monkey to study the functional organization of curvature representation. We find a modular basis for cue-invariant curvature representation in area V4 of monkey visual cortex and illustrate a systematic representation from low to high curvature and of curvature orientation, replete with curvature pinwheels. This is the first report of systematic functional organization for curvature representation in the visual system. The use of optical imaging has revealed at a population level spatial details of cortical responses, something which has not been evident from previous studies of single neurons. These data support a representational architecture underlying a ‘curvature hypercolumn’ in V4.

Comparison of stimulus-evoked BOLD responses in human and monkey visual cortex

We characterized the blood oxygenation level dependent (BOLD) signal in humans and macaque monkeys by comparing the response in visual cortex to a single checkerboard or two checkerboards, spaced 1.5, 3.0, or 4.5 s apart. We found that the magnitude and shape of the BOLD response to a single checkerboard was similar in the two species. In addition, we found that the BOLD responses summed similarly, and that at an inter-stimulus interval (ISI) of 4.5 sec BOLD summation was nearly linear in both species. When comparing the ratio of the amplitude of the response to the second checkerboard at the 4.5 sec ISI with that of the single checkerboard between subjects in both species, the results from both monkey subjects fell within one standard deviation of the mean human results (human mean (n=12): .95 +/− .31 second/single response amplitude; monkey 1: 1.16; monkey 2: .86). At the shorter ISIs, both species demonstrated increased suppression of the BOLD response to the second checkerboard. These findings indicate that the magnitude of the BOLD response to events separated by 4.5 seconds can be accurately measured in and compared between human and monkey visual cortex.

Human cortical and behavioral sensitivity to patterns of complex motion at eccentricity

Complex patterns of image motion (contracting, expanding, rotating, and spiraling fields) are important in the coordination of visually guided behaviors. Whereas specialized detectors in monkey visual cortex show selectivity for particular patterns of complex motion, their representation in human visual cortex remains unclear. In the present study, functional magnetic resonance imaging (fMRI) was used to investigate the sensitivity of functionally defined regions of human visual cortex to parametrically modulated complex motion trajectories, coupled with complementary psychophysical testing. A unique stimulus design made it possible to disambiguate the neural responses and psychophysical sensitivity to complex motions per se from the distribution of local motions relative to the fovea, which are known to enhance cortical activity when presented radial to fixation. This involved presenting several small, separate motion fields in the periphery in a manner that distinguished them from global optic flow patterns. The patterns were morphed through complex motion space in a systematic time-locked fashion when presented in the scanner. Anisotropies were observed in the fMRI signal, marked by an enhanced response to expanding vs. contracting fields, even in early visual cortex. Anisotropies in the psychophysical sensitivity measures followed a similar pattern that was correlated with activity in areas hV4, V5/MT, and MST. This represents the first systematic examination of complex motion perception at both a behavioral and neural level in human observers. The characteristic processing anisotropy revealed in both data sets can inform models of complex motion processing, particularly with respect to computations performed in early visual cortex.