Encoding of graded changes in spatial specificity of prior cues in human visual cortex

Abstract

Prior information about the relevance of spatial locations can vary in specificity; a single location, a subset of locations, or all locations may be of potential importance. Using a contrast-discrimination task with four possible targets, we asked whether performance benefits are graded with the spatial specificity of a prior cue and whether we could quantitatively account for behavioral performance with cortical activity changes measured by blood oxygenation level-dependent (BOLD) imaging. Thus we changed the prior probability that each location contained the target from 100 to 50 to 25% by cueing in advance 1, 2, or 4 of the possible locations. We found that behavioral performance (discrimination thresholds) improved in a graded fashion with spatial specificity. However, concurrently measured cortical responses from retinotopically defined visual areas were not strictly graded; response magnitude decreased when all 4 locations were cued (25% prior probability) relative to the 100 and 50% prior probability conditions, but no significant difference in response magnitude was found between the 100 and 50% prior probability conditions for either cued or uncued locations. Also, although cueing locations increased responses relative to noncueing, this cue sensitivity was not graded with prior probability. Furthermore, contrast sensitivity of cortical responses, which could improve contrast discrimination performance, was not graded. Instead, an efficient-selection model showed that even if sensory responses do not strictly scale with prior probability, selection of sensory responses by weighting larger responses more can result in graded behavioral performance benefits with increasing spatial specificity of prior information.

Two different mechanisms support selective attention at different phases of training

“…The behavioral data presented here were divided into 2 training phases in order to match the data from the main EEG analysis, in which we tried to obtain ~400 trials for each experimental condition (10 EEG sessions for each training phase) as suggested by Luck [54]. In line with previous studies, Δc increased as a function of contrast (F(5, 55) = 143.38, p < 0.001) [9,28,52,55–58] and decreased with focused compared to divided attention (F(1, 11) = 137.95, p < 0.001) [5,9,28,56]. These attentional modulations occurred at all contrast levels (all t(11)’s ≥ 5.03, all p ’s < 0.001, data collapsed across early and late training phases, false discovery rate [FDR]-corrected, 2-tailed).…”

“…We used this task to make contact with previous studies in both humans and nonhuman primates that have employed similar paradigms [5,9,36–38]. At the start of a trial, subjects were cued to attend to either the left or right lower visual quadrant (termed focused attention trials), or they were cued to attend to both locations (termed divided attention trials).…”

“…The main dependent measure was the change in contrast (i.e., contrast threshold or Δ c ) from each pedestal contrast that was required to achieve an accuracy level of 76% (perceptual sensitivity or d’ of approximately 1). When paired with simultaneous EEG recording, this psychophysical approach allowed us to derive both psychometric and neurometric response functions so that we could directly link attentional modulations in neural responses with attentional modulations in behavior (see also [5,9,28,47,51–53]).…”

“…For example, recent studies propose that noise reduction or efficient decoding schemes can facilitate selective processing even in the absence of gain [5,28,36–37,40,43–44,51]. To evaluate these alternative accounts, we adopted a quantitative modeling framework based on SDT to evaluate the impact of gain and noise modulations on behavioral performance [5,9,28,47,51–53]. Later, we also evaluated possible contributions of efficient decoding [5,28].…”

“…However, later in training, there was still an attentional facilitation of behavioral performance even though attentional gain of the P1 component vanished. Next, we adopted a standard quantitative model, based on signal detection theory (SDT), to infer whether attentional gain or noise reduction best predicted the relationship between attentional modulations of the P1 component and behavior across training sessions [5,9,28,47,51–53]. Over the course of training, the SDT model suggests a link between P1 gain and behavioral gain early in training but noise reduction later in training.…”