Anna Sterkin

Training the brain to overcome the effect of aging on the human eye

Presbyopia, from the Greek for aging eye, is, like death and taxes, inevitable. Presbyopia causes near vision to degrade with age, affecting virtually everyone over the age of 50. Presbyopia has multiple negative effects on the quality of vision and the quality of life, due to limitations on daily activities – in particular, reading. In addition presbyopia results in reduced near visual acuity, reduced contrast sensitivity, and slower processing speed. Currently available solutions, such as optical corrections, are not ideal for all daily activities. Here we show that perceptual learning (repeated practice on a demanding visual task) results in improved visual performance in presbyopes, enabling them to overcome and/or delay some of the disabilities imposed by the aging eye. This improvement was achieved without changing the optical characteristics of the eye. The results suggest that the aging brain retains enough plasticity to overcome the natural biological deterioration with age.

ADHD subjects fail to suppress eye blinks and microsaccades while anticipating visual stimuli but recover with medication.

Oculomotor behavior and parameters are known to be affected by the allocation of attention and could potentially be used to investigate attention disorders. We explored the oculomotor markers of Attention-deficit/hyperactivity disorder (ADHD) that are involuntary and quantitative and that could be used to reveal the core-affected mechanisms, as well as be used for differential diagnosis. We recorded eye movements in a group of 22 ADHD-diagnosed patients with and without medication (methylphenidate) and in 22 control observers while performing the test of variables of attention (t.o.v.a.). We found that the average microsaccade and blink rates were higher in the ADHD group, especially in the time interval around stimulus onset. These rates increased monotonically over session time for both groups, but with significantly faster increments in the unmedicated ADHD group. With medication, the level and time course of the microsaccade rate were fully normalized to the control level, regardless of the time interval within trials. In contrast, the pupil diameter decreased over time within sessions and significantly increased above the control level with medication. We interpreted the suppression of microsaccades and eye blinks around the stimulus onset as reflecting a temporal anticipation mechanism for the transient allocation of attention, and their overall rates as inversely reflecting the level of arousal. We suggest that ADHD subjects fail to maintain sufficient levels of arousal during a simple and prolonged task, which limits their ability to dynamically allocate attention while anticipating visual stimuli. This impairment normalizes with medication and its oculomotor quantification could potentially be used for differential diagnosis.

Learning to adapt: Dynamics of readaptation to geometrical distortions

The visual system can adapt to optical blur, whereby the adapted image is perceived as sharp. Here we show that adaptation reduces blur-induced biases in shape perception, with repeated adaptations (perceptual learning), leading to unbiased perception upon re-exposure to blur. Observers wore a cylindrical lens of +1.00 D on one eye, thus simulating monocular astigmatism. The other eye was either masked with a translucent blurred lens (monocular) or unmasked (dichoptic). Adaptation was tested in several repeated sessions with a proximity-grouping task, using horizontally or vertically arranged dot-arrays, without feedback, before, after, and throughout the adaptation period. A robust bias in global-orientation judgment was observed with the lens, in accordance with the blur axes. After the observer wore the lens for 2 h, there was no significant change in the bias, but after 4 h, the monocular condition, but not the dichoptic, resulted in reduced bias. The adaptation effect of the monocular 4-h adaptation was preserved, and even improved, when the lens was re-applied the next day, indicating learning. After-effects were observed under all experimental conditions except for the 4-h monocular condition, where learning took place. We suggest that, with long experience, adaptation is transferred to a long-term memory that can be instantly engaged when blur is re-applied, or disengaged when blur is removed, thus leaving no after-effects. The comparison between the monocular and dichoptic conditions indicates a binocular cortical site of plasticity.

Spatio-temporal low-level neural networks account for visual masking

T emporal masking is a paradigm that is widely used to study visual information processing. When a mask is presented, typically within less than 100 msec before or after the target, the response to the target is reduced. The results of our psychophysical and visual evoked potential (VEP) experiments show that the masking effect critically depends on a combination of several factors: (1) the processing time of the target, (2) the order of presentation of the target and the mask, and (3) the spatial arrangement of the target and the mask. Thus, the masking effect depends on the spatial-temporal combination of these factors. Suppression was observed when the mask was positioned within a spatial range that was found to evoke inhibition, and when the temporal separation between the target and the mask was short. In contrast, lateral facilitation was observed when the mask was presented at a spatial separation that did not evoke inhibition from the target’s vicinity and with a temporal sequence that preceded the target, or when it was presented simultaneously with it, but not when the target preceded the mask. We propose that masking effects, either suppression or facilitation, reflect integration into the spatial and the temporal domains of the feedforward response to the target and the lateral inputs evoked by the mask (excitatory and/or inhibitory). Because the excitation evoked by the mask develops and propagates slowly from the mask’s location to the target’s location, it lags behind the response to the target. On the other hand, inhibition that is produced in the vicinity of the target evolves more rapidly and follows the onset and offset of the stimulus more closely. Thus, lateral excitation that overcomes the inhibition may facilitate the grouping of local elements into a global percept by increasing the survivability of the object and its accessibility for perceptual awareness.

Backward masking suppresses collinear facilitation in the visual cortex.

Perceptual facilitation in detecting low-contrast Gabor patches (GPs) is induced by collinearly oriented high-contrast flankers. Our recent Visual Evoked Potentials (VEPs) study provided new physiological evidence for these collinear interactions, reflected by nonlinear modulation of multiple waveform components and frequencies [Sterkin, A., Yehezkel, O., Bonneh, Y. S., Norcia, A., & Polat, U. (2008). Multi-component correlate for lateral collinear interactions in the human visual cortex. Vision Research, 48(15), 1641-1647]. Here we used VEPs to study the temporal structure of this process. Low-contrast, foveal target GP (T) was simultaneously flanked by two collinear high-contrast GPs with a spatial separation that induces facilitation of T (lateral masking, LM). Another mask, identical to LM, was presented at different time-intervals (ISIs) after LM (backward masking, BM-on-LM). The responses were compared to separate waveforms evoked by T-alone and mask-alone at different ISIs. BM canceled the physiological markers of facilitation at an ISI of 50 ms, in agreement with earlier psychophysical findings, whereas no BM effect on T-alone was observed. This ISI coincides with the active time-window of lateral interactions, confirming our working model. The waveform amplitude of the negative N1 peak of LM was modulated toward the linear prediction of no interactions and the spectrum was shifted toward suppression, with no evidence of facilitation. Moreover, the P1 peak amplitude of BM was decreased at the same ISI, indicating that there is a mutual interference in cortical representation of both events. Waveform subtraction between BM-on-LM and LM suggests a mechanism of extended persistence of the target representation underlying facilitation in LM. We suggest an explanation for the role of improved detection of collinear stimuli in grouping of contours.

Response similarity as a basis for perceptual binding

Detection of low-contrast Gabor patches (GPs) is improved when flanked by collinear GPs, whereas suppression is observed for high-contrast GPs. The facilitation resembles the principles of Gestalt theory of perceptual organization. We propose a model for contour integration in the context of noise that incorporates a temporal element into this spatial architecture. The basic principles are (1) the response increases with increasing contrast, whereas the latency decreases; (2) activity-dependent interactions: facilitation for low and suppression for high activity; (3) the variance increases with contrast for responses, rates, and latency; and (4) inhibition has a shorter time constant than excitation. When a texture of randomly oriented GPs is presented, the response to every element decreases due to fast inhibition between the neighboring elements, shifting the activity toward the range of collinear facilitation. Next, the slower excitation induces selective facilitation along the contour elements. Consequently, the response to the contour increases, whereas the variance of the rate and latency decreases, providing better temporal correlation between the contour elements. Thus, collinear facilitation increases the saliency of contours. Our model may suggest a solution to the binding problem by bridging between the temporal and spatial aspects of lateral interactions that determine the encoding of perceptual grouping.

Multi-component correlate for lateral collinear interactions in the human visual cortex

Perceptual facilitation, a decrease in detection threshold for low-contrast Gabor patches (GPs) occurs when the GP is flanked by collinearly oriented high-contrast patches. There is earlier evidence suggesting a spatial architecture of excitatory and inhibitory interactions. Here we used Visual Evoked Potentials (VEPs) to study the temporal structure of this process. We measured VEPs elicited by a foveal near-threshold target GP presented in isolation (T), T in the presence of two flanking collinear high-contrast GPs (lateral masking, LM), or the flankers alone (F). Stimuli were presented for 50 ms every 1000 ms. The choice of the set parameters elicited behavioral facilitation of T detection. Significant modulation of peak amplitudes in LM compared with linearly summed waveforms elicited by T and F was found for five alternating polarity components, ranging from 65 to 290 ms after stimulus onset. In the frequency domain, suppression at lower frequencies (up to 0.8 log units) was followed by facilitation at higher frequencies (4-6 Hz, up to 0.8 log units). Although no differences in the latencies were found, lateral interactions were reflected by non-linear waveform modulation of multiple components and frequencies, including components as early as 65-75 ms. Spectrum analysis suggests that both suppression and facilitation may be found for the same configuration of stimuli, simultaneously, distributed at different temporal frequencies and/or sources. The physiological correlates of lateral interactions may thus originate at multiple sources, only some of which are explicitly facilitatory. The final perceptual outcome of this complex spatio-temporal representation is determined by combining sensory and cognitive factors.

Gains following perceptual learning are closely linked to the initial visual acuity

The goal of the present study was to evaluate the dependence of perceptual learning gains on initial visual acuity (VA), in a large sample of subjects with a wide range of VAs. A large sample of normally sighted and presbyopic subjects (N = 119; aged 40 to 63) with a wide range of uncorrected near visual acuities (VA, −0.12 to 0.8 LogMAR), underwent perceptual learning. Training consisted of detecting briefly presented Gabor stimuli under spatial and temporal masking conditions. Consistent with previous findings, perceptual learning induced a significant improvement in near VA and reading speed under conditions of limited exposure duration. Our results show that the improvements in VA and reading speed observed following perceptual learning are closely linked to the initial VA, with only a minor fraction of the observed improvement that may be attributed to the additional sessions performed by those with the worse VA.

Binocular combination of stimulus orientation

When two sine waves that differ slightly in orientation are presented to the two eyes separately, a single cyclopean sine wave is perceived. However, it is unclear how the brain calculates its orientation. Here, we used a signal detection rating method to estimate the perceived orientation when the two eyes were presented with Gabor patches that differed in both orientation and contrast. We found a nearly linear combination of orientation when both targets had the same contrast. However, the binocular percept shifted away from the linear prediction towards the orientation with the higher contrast, depending on both the base contrast and the contrast ratio. We found that stimuli that differ slightly in orientation are combined into a single percept, similarly for monocular and binocular presentation, with a bias that depends on the interocular contrast ratio. Our results are well fitted by gain-control models, and are consistent with a previous study that favoured the DSKL model that successfully predicts binocular phase and contrast combination and binocular contrast discrimination. In this model, the departures from linearity may be explained on the basis of mutual suppression and mutual enhancement, both of which are stronger under dichoptic than monocular conditions.

Increased gamma band activity for lateral interactions in humans

Collinear facilitation of contrast sensitivity supported by lateral interactions within primary visual cortex is implicated in contour and object perception, with neural correlates in several frequency bands. Although higher component of the ERP power spectrum, the gamma-band, is postulated to reflect object representation, attention and memory, its neuronal source has been questioned, suggesting it is an artifact reflecting saccadic eye movements. Here we explored the gamma-band activity during collinear facilitation with no saccade-related confounds. We used single-trial spectral analysis of ERP in occipital channels in a time-window of nearly complete saccadic suppression and discarded sporadic trials containing saccades, in order to avoid saccadic artifacts. Although converging evidence suggests that gamma-band oscillations emerge from local excitatory–inhibitory balance involving GABAergic inhibition, here we show activity amplification during facilitatory collinear interactions, presumably dominated by excitations, in the gamma-band 150–350 milliseconds following onset of low near-threshold contrast stimulus. This result highlights the potential role of gamma-band oscillations in neuronal encoding of basic processes in visual perception. Thus, our findings suggest that gamma-band ERP spectrum analysis may serve as a useful and reliable tool for exploring basic perception, both in normal adults and in special populations.