Priming of Attentional Selection in Macaque Visual Cortex: Feature-Based Facilitation and Location-Based Inhibition of Return

Abstract

Visual search performance varies with stimulus and response history. Priming of pop-out refers to increased accuracy and reduced response time with repeated presentation of particular singleton and distractor features (e.g., a red target among green distractor stimuli), which are abruptly impaired when singleton and distractor features swap (e.g., green target among red distractors). Meanwhile, inhibition of return refers to slowing of response time when target location repeats. Neurophysiological correlates of both these phenomena have been reported in the frontal eye field (FEF), an area in the frontal lobe contributing to attentional selection and eye movement planning. To understand the mechanistic origin of these adaptive behaviors, we investigated visual cortical area V4, an area providing input to and receiving feedback from FEF, during feature-based priming of popout and location-based inhibition of return. Performing a color pop-out task, monkeys exhibited pronounced priming of pop-out and inhibition of return. Neural spiking from V4 revealed earlier target selection associated with priming of pop-out and delayed selection associated with inhibition of return. These results demonstrate substantial involvement of extrastriate visual cortex in behavioral priming and inhibition of return. Significance Statement Mid-level attention and visual processing is influenced by recent history of visual stimuli and gaze behavior. Using priming of pop-out visual search, we discovered that neural spiking in extrastriate visual area V4 shows speeded attentional selection when target and distractor features repeat and delayed selection when target location repeats. These neural processes paralleled but did not account for the magnitude of visual search performance changes with stimulus and response history. These new results improve our understanding of how recent experience influences attention and performance.

Gravitational models explain shifts on human visual attention

“…This mechanism is called Inhibition Of Return (IOR). Originally discovered in human studies of attention, inhibition of return is a tendency for the organism to orient away from a previously attended location and biologically depends on neural structures that participate in oculomotor control 29 , parietal and frontal cortex 37 . This phenomenon was first described by Posner and Cohen 38 who showed that reaction times to detect objects appearing in previously cued locations were longer than to uncued locations.…”

“…In the proposed framework, the encoding of the oculomotor command needs a simple neuronal hardware, with the very mild assumption of units that perform sum operation, without the need of backward flows, leading to a plausible and straightforward biological implementation. The literature indicates that the V1 area is important for the conformation of the bottom-up saliency 27 , 28 , while other associative areas such as V4, FEF and supplementary eye field (SEF) 29 receive signals simultaneously from both the V1 area and deeper layers 30 . Neurons in the V1 area can encode principal and independent components extracted directly from the visual input 28 , and the response magnitude of such neurons is greater when the stimulus is distinct from its surrounding 27 , 28 .…”

Salience by Competitive and Recurrent Interactions: Bridging Neural Spiking and Computation in Visual Attention

“…In target search, the comparison is between objects in the display and a target representation; in popout search, the comparison is between objects in the display and one another. In both cases, the representations depend on ventral-stream visual areas like V4 and IT (Zhou & Desimone, 2011;Westerberg, Maier, & Schall, 2020) as well as prefrontal areas (Bichot et al, 2015;Bichot, Xu, Ghadooshahy, Williams, & Desimone, 2019). The processing involved in popout is likely related to the identification-based feedforward inhibition we already identified as important for target selection.…”

Stimulus Feature-Specific Information Flow Along the Columnar Cortical Microcircuit Revealed by Multivariate Laminar Spiking Analysis

“…All procedures were in compliance with regulations set forth by the Association for the Assessment and Accreditation of Laboratory Animal Care (AALAC), approved by the Vanderbilt University Institutional Animal Care and Use Committee, and followed National Institutes of Health guidelines. A detailed description of the surgical procedures can be found in previous publications (Westerberg et al, 2019(Westerberg et al, , 2020a. Briefly, in a series of surgeries, each monkey was implanted with a custom MRI-compatible headholder and recording chamber over perifoveal V1 concurrent with a craniotomy.…”

“…Channels in the gray matter were found by determining first whether a visual response could be evoked on the channel and second, whether a receptive field was present for the multiunit and/or LFP activity through a previously described receptive field mapping paradigm (Westerberg et al, 2019). If the channel was found to be in the gray matter, the broadband neural signal recorded at that channel was then band-pass filtered between 500 and 5,000 Hz, rectified, and low-pass filtered at 200 Hz using Butterworth filters (Self et al, 2013;Shapcott et al, 2016;Westerberg et al, 2020a). These derived neural signals, with no further filtering of the multiunit activity, were then used in performing both the univariate and multivariate analyses ( Figure 1D).…”