Attention allocation & visual processing

• Visual processing in attention may be regarded as a limited resource due to the brain's constraints in handling multiple stimuli simultaneously, impacting overall performance.
• Enhancing an attended stimulus should invariably lead to the suppression of unattended stimuli.
• Feature-selective attention shifts were examined through voluntary cues on superimposed red or blue random dot kinematograms (RDKs) to test whether there is a reciprocal relationship between the attended enhancement and unattended suppression.
• The steady-state visual evoked potential (SSVEP), an oscillatory brain response from flickering RDKs, was measured in human EEG.
• Supporting limited resources, both attended enhancement and unattended suppression were observed although not concurrently.
• Attended RDK enhancement started at 220 ms post-cue onset and preceded unattended RDK suppression by about 130 ms.
• This significant correlation suggests that the fluctuations observed in the SSVEP signal patterns over time have a direct impact on the individual's behavioral responses. Such variability can reflect how effectively the brain processes and reacts to visual information, highlighting the dynamic interplay between sensory perception and attentional focus.
• Significant deviations from temporally synchronized reciprocity between enhancement and suppression suggested that enhancement caused suppression.
• Shifting and focusing attention on a certain location, object, or feature is a key element in sensory information extraction for adaptive behavior.
• The distribution of attentional resources and temporal neural mechanisms of attentional shifting are not well understood.
• Prior studies measured event-related potentials (ERPs) during the cue-target interval which investigates neural mechanisms of cue processing and target expectation, and cortical control network activity.
• ERP’s cannot provide information on the temporal dynamics of neural facilitation with suppression in relevant early processing areas that are involved in processing a new stimulus
• Behavioral performance is closely linked to cortical evoked activity modulation.
• Visual processing as a strictly limited resource suggests attended enhancement should be accompanied by an equal suppression of the unattended stimulus.
• Shifting attention to one of two lateral flickering stimuli was purely facilitatory without significant suppression of the unattended-location stimulus in the opposite visual field in previous studies.
• Attention provides additional resources rather than influencing a distribution of limited resources.
• Attentional resources may not be shared across visual hemifields.
• Suppressive stimulus interactions were found when stimuli within the same hemifield were less than 4° of visual angle apart.
• Mutual suppression between multiple stimuli is greatest when they fall into the same receptive fields.
• Attentional modulation is mainly facilitatory, even when attended and unattended stimuli occupy the same spatial location.
• Two completely overlapping random dot kinematograms (RDKs) of different colors (red or blue) were presented to estimate the speed and neural dynamics in the visual cortex during shifts of feature-selective attention.
• Participants monitored the cued color's RDK to detect brief coherent motion intervals (targets) while ignoring distractors in the unattended RDK.
• Both RDKs flickered at different rates, eliciting distinguishable steady-state visual evoked potentials (SSVEPs).
• The SSVEP is an ongoing oscillatory response from a flickering stimulus indexing neural activity responding continuously.
• The SSVEP has the same temporal frequency as the driving stimulus, and its amplitude is enhanced by both spatial and feature-selective attention.
• Allocation of processing resources to both the attended and the unattended stimulus were concurrently measured.
• Any enhancement of SSVEP amplitudes of the attended stimulus should be accompanied by a mirrored suppression of the unattended stimulus because visual processing at a shared spatial location is a strictly limited resource; i.e., the sum of both signals should remain constant.

Results

• Participants were able to discriminate targets from distractors indicated by the average observer.
• Reaction times to targets as a function of the cue-target interval became rapidly faster until the third time bin (287–429 ms).
• Paired t tests between successive time bins were significant from the first (0–143 ms) to the second (144–286 ms) time bin and from the second to the third time bin.
• None of the comparisons between later time bins were significant.
• SSVEP amplitudes averaged across attentional conditions showed a narrow peak at occipital electrodes for both RDKs

Steady State Visual Evoked Potential

• SSVEP amplitude time courses for the to-be-attended RDK exhibited a long-lasting amplitude facilitation.
• The to-be-ignored RDK showed an amplitude suppression that started later but lasted throughout the stimulation period.
• The first time point having a significant amplitude increase for the attended stimulus was at 223 ms, lasting until 1,059 ms after cue onset
• The first significant time point with amplitude reduction for the to-be-ignored stimulus was about 130 ms later at 356 ms after cue onset, which remained significant until the analyzed time range's end.
• Significance in enhancement and suppression were tested by calculating measures from the observed courses of SSVEP amplitudes termed selectivity and total activity.
• Both time courses of measures showed significant changes over time
• Selectivity became significantly larger than zero from 242 ms postcue until the analyzed time range's end.
• Total activity showed a significant increase from 266 to 520 ms postcue and a significant reduction from 785 ms postcue until the end of the analyzed time range.
• Correlations were found between reaction times as a function of cue-target interval within the initial 800 ms after cue onset
• Higher amplitudes related to faster responses for the attended stimulus, as lower amplitudes related to faster responses for the ignored stimulus.
• Higher values of selectivity were related to faster responses.
• Total activity showed no significant correlation with reaction times.
• The maximum attentional modulations occurred in the posterior medial occipital cortex, which contains the early visual areas V1, V2, and V3.

Discussion

• Steady-state stimulation technique was uniquely used to investigate temporal neural dynamics of feature-selective attentional shifting in processing areas.
• Enhancement of the attended and suppression of the unattended stimulus were observed using spatially overlapping stimuli.
• Shifting attention to either color of two spatially overlapping RDKs was linked to a bi-phasic process.
• Beginning around 220 ms after color cue onset, the SSVEP amplitude to the to-be-attended RDK was significantly increased
• A significant amplitude reduction occurred at the to-be-ignored RDK roughly 360 ms after the cue.
• The cortical currents giving rise to the SSVEP attention effect were localized to a region containing the early visual areas V1-V3, supporting previous studies on these regions
• Maximum behavioral performance was reflected in fastest reaction times which was reached in a time bin from 286 to 429 ms after the cue.
• There were significant correlations between behavioral data and amplitude augmentation with suppression within the first 800 ms after cue onset.
• Total activity was not correlated with the time course of reaction times
• Selectivity showed significant correlations with the temporal changes in reaction times.
• Behavioral performance was best explained by the relative difference in processing between attended and unattended stimuli.
• Enhancement and suppression during attentional shifts was measured relative to a precue baseline.
• Participants directed attention to either the red or the blue RDK during the baseline interval before the cue.
• An arbitrary selection during baseline would lead to higher SSVEP amplitudes driven by the red and to lower amplitudes for the blue RDK in some trials and vice versa in others, thereby causing a negative correlation of SSVEP amplitudes over trials.
• A recent study demonstrates that an increase of firing rates to an attended stimulus was followed by a reduction of firing rates to an unattended stimulus.
• Attention to a feature can suppress responses of neurons across the visual field tuned to the opposite feature value.
• Enhancement with suppression was linearly related to the difference in degrees between the attended motion direction and the neuron's preferred direction,
• Temporal dissociation suggests that attentional selection of color differs from attentional selection of motion direction, or the colors presented here are orthogonal in respect to attentional selection.
• The biphasic time course indicates two different neural mechanisms
• Sensory gain mechanism enhances attended stimuli
• Suppressive effects of multiple concurrently presented stimuli are within the framework of the "biased competition model".
• Enhancement of the attended stimulus in early visual areas with small receptive fields would propagate to later areas in the visual stream.
• The strong competition between both stimuli would lead to the suppression of the unattended stimulus.
• Trials with stronger suppression also showed stronger enhancement before, but not during or after suppression compared with trials with weak suppression
• Tracking of individual dots, surface-based attention, or attentional selection of flicker frequencies rather than color cannot explain results alternatively.
• Earliest modulations of processing are stimulus processing in the form of a selection negativity beginning around 170 ms post stimulus onset and doesn't involve interpreting to shift attention
• Modulations seemed too early because previous experiments had the to-be-attended colors known before stimulus onset whereas the present study had a modulation onset at 220ms postcue
• Recent findings confirm this rapid processing due to feature-selective attention can modulate processing earlier if resources are directly competed for.
• Cospatial color stimuli in feature-selective attentional shifting showed a previously undescribed resource time course: The enhancement of an attended feature and a delayed suppression of the unattended feature