Skip to main content
NCBI home page
Cerebral Cortex (New York, NY) logoLink to Cerebral Cortex (New York, NY)
. 2020 Feb 6;30(5):2823–2833. doi: 10.1093/cercor/bhz277

Impaired Fixation-Related Theta Modulation Predicts Reduced Visual Span and Guided Search Deficits in Schizophrenia

Elisa C Dias 1,2,✉, Abraham C Van Voorhis 1, Filipe Braga 1,2, Julianne Todd 1, Javier Lopez-Calderon 1,3, Antigona Martinez 1,3, Daniel C Javitt 1,3,✉
PMCID: PMC7197077  PMID: 32030407

Abstract

During normal visual behavior, individuals scan the environment through a series of saccades and fixations. At each fixation, the phase of ongoing rhythmic neural oscillations is reset, thereby increasing efficiency of subsequent visual processing. This phase-reset is reflected in the generation of a fixation-related potential (FRP). Here, we evaluate the integrity of theta phase-reset/FRP generation and Guided Visual Search task in schizophrenia. Subjects performed serial and parallel versions of the task. An initial study (15 healthy controls (HC)/15 schizophrenia patients (SCZ)) investigated behavioral performance parametrically across stimulus features and set-sizes. A subsequent study (25-HC/25-SCZ) evaluated integrity of search-related FRP generation relative to search performance and evaluated visual span size as an index of parafoveal processing. Search times were significantly increased for patients versus controls across all conditions. Furthermore, significantly, deficits were observed for fixation-related theta phase-reset across conditions, that fully predicted impaired reduced visual span and search performance and correlated with impaired visual components of neurocognitive processing. By contrast, overall search strategy was similar between groups. Deficits in theta phase-reset mechanisms are increasingly documented across sensory modalities in schizophrenia. Here, we demonstrate that deficits in fixation-related theta phase-reset during naturalistic visual processing underlie impaired efficiency of early visual function in schizophrenia.

Keywords: eye movements, fixation-related potential (FRP), intertrial coherence, visual search

Introduction

Schizophrenia is associated with deficits in early sensory processing that contribute to higher-order disabilities (reviewed in Javitt (2015) and Javitt and Freedman (2015)). In the visual system, patients show consistent deficits in magnocellular function as demonstrated using behavioral, functional magnetic resonance imaging, and neurophysiological approaches (Martinez et al. 2012; Bedwell et al. 2013; Gracitelli et al. 2013; Leonard et al. 2014; Kim et al. 2015; Martinez et al. 2015). These deficits, moreover, interrelate with impairments in higher-order visual functions including perceptual closure, visual working memory, and face emotion recognition (Doniger et al. 2002; Butler et al. 2009; Dias et al. 2011).

To date, the vast majority of physiological studies of visual processing in schizophrenia have utilized stimuli presented while subjects fixate on a single spot (“central fixation”). By contrast, in natural vision, individuals continually move their eyes to seek out relevant information through a series of saccades and fixations, a process termed “active sensing” (Konig and Luksch 1998; Schroeder et al. 2010; Ito et al. 2013; Hessels et al. 2018; Concha-Miranda et al. 2019). Integrity of these processes has been studied in schizophrenia to only a limited degree. Here, we utilize eye-tracking combined with neurophysiological assessment to evaluate both causes and consequences of active sensing deficits in schizophrenia during a visual Guided-Search task.

In Guided Search, subjects view successive search-fields containing arrays of multiple background stimuli (“distractors”) as well as a single predesignated “target” stimulus (Treisman 1982; Nuechterlein et al. 2009). To find the target, subjects must sample the search-field through a series of saccades and fixations. The time required to complete a search-field depends primarily upon the number of eye movements needed. At each fixation, individuals process visual information with high spatial resolution only within the “foveal” region, that is, approximately the central ∼1° of visual space. By contrast, the surrounding “parafoveal” (∼1–5°) and “perifoveal” (∼5–10°of visual space) are sensitive primarily to low spatial frequency information, with a fall-off of sensitivity with increasing eccentricity (Poletti et al. 2017).

When targets differ from distractors in only a single physical feature (“feature-search”), target stimuli “pop out” from the visual field, permitting “efficient”, or “parallel-search” of the field (Treisman 1982; Nuechterlein et al. 2009). By contrast, when targets differ by a conjunction of features (“conjunction-search”), subjects must use top–down attentional allocation to serially “shine a spotlight” on subsets of stimuli for further detailed evaluation (Treisman 1982). In this case, the search time increases linearly with the number of stimuli (set-size) within the stimulus array even as the size of the search-field remains constant.

The slope of the increase in reaction time (RT) with set-size indicates the degree to which subjects successfully limit their search only to a subset of stimuli, and thus may serve as an index of integrity of top–down attentional control (Nuechterlein et al. 2009). It has been shown that distinct networks of cortical areas are activated for the different types of search due to the different strategies required in each. (Nobre et al. 2003).

When subjects fixate, they generate an occipital fixation-related potential (FRP)—also termed the λ potential (Fourment et al. 1976; Kazai and Yagi 2003). As opposed to more traditional visual event-related potentials (ERP) that are time-locked to the presentation of a stimulus, FRPs are time-locked to fixation onset as detected using an external eye tracker (Kamienkowski et al. 2012). Although FRPs are sensitive to stimulus properties in the to-be-fixated region, they occur even in the dark, or while subjects are viewing a blank screen, and are thought to reflect “priming” of visual cortex to process the to-be-fixated information rather than response to the stimulus itself (Rajkai et al. 2008; Ito et al. 2013).

In nonhuman primates, the FRP consists primarily of a phase-concentration in the theta- (4–8 Hz) frequency range (Rajkai et al. 2008). Visual search tasks are especially well suited to study FRPs in active sensing as the subject is required to fixate on a series of similar stimuli while they search for the target, thus providing many fixations per trial.

Here, we used distractor stimuli that differed in either orientation-alone (feature-search) or a conjunction of both contrast and orientation (conjunction-search) (Fig. 1A). In the feature (parallel) search condition, all stimuli were low-contrast, and target stimuli differed only in orientation (vertical vs. horizontal). In the conjunction (serial) search condition, the target was a single vertically oriented, low-contrast stimulus, among more frequent high-contrast vertical and low-contrast horizontal stimuli. Based upon the stimulus configurations, we hypothesized that search times in the feature-search condition would be independent of set-size within the array, whereas in the conjunction-search condition the RT would increase linearly with increases in the number of stimuli.

Figure 1.

Figure 1

Open in a new tab

(A) Example search paths from a representative control and patient with performance similar to the group mean. Circles represent random-search fixations, with diameter representing fixation duration. Gray squares represent prefinal fixation. Thick black circles show final fixation on target. Dashed circle indicates location of the target stimulus. (B) Pilot Behavior Study results. Mean (SEM) search time per search-field as a function of set-size during serial search (note log RT scale). In rmANOVA across conditions by subject, a significant effect of group (P = 0.001) and a significant linear effect of density (P < 0.0001) were observed the group X density effect was nonsignificant (P = 0.7) **P < 0.01 across densities.

Eye-tracking data were collected along with target detection time to permit with analyses across groups as well as calculation of the FRPs. In an initial behavior pilot study (PS), we evaluated behavioral performance across a range of stimulus contrasts and array densities. In a subsequent neurophysiological study, we chose a single contrast variation and set-size and collected FRPs along with behavioral measures. We hypothesized that FRP generation would be significantly reduced in schizophrenia, and that deficits would correlate with impaired search-related neurocognitive impairments.

Materials and Methods

Subjects

Subjects (Table 1) were recruited from the Nathan Kline Institute and affiliated clinics, and signed informed consent following full description of study procedures. The study was approved by the Institutional Review Board associated with the NYS Office of Mental Health.

Table 1.

Demographic, neurocognitive and visual search parameters

Pilot Study Neurophysiology Study
Controls (HC) Patients (SzP) Controls (HC) Patients (SzP)
(N = 15) (N = 15) (N = 25) (N = 24)
Demographics
 Age 35.9 (10.0) 34.5 (9.0) 35.6 (9.3) 36.6 (9.6)
 Parental SESa 39.9 (13.1) 36.1 (13.5) 42.0 (14.6) 38.4 (12.3)
 Individual SESb 40.7 (11.8) 23.0 (10.4)** 43.9 (11.9) 24.9 (7.7)**
 Years educationc 14.4 (3.0) 11.4 (1.6)** 15.3 (2.1) 12.0 (2.4)**
 Quick IQb 103.7 (11.5) 94.4 (13.6) 105.6 (8.7) 97.3 (8.5)**
MCCBd
 Processing speede 49.5 (9.5) 29.0 (10.5)** 51.5 (10.3) 33.3 (11.8)**
 Attention/vigilancef 51.1 (5.9) 32.4 (12.7)** 52.1 (10.8) 36.2 (11.7)**
 Working memorye 47.5 (11.3) 34.7 (11.2)* 50.9 (9.5) 38.0 (10.1)**
 Verbal learninge 44.0 (6.5) 36.7 (8.2) 47.4 (8.6) 36.2 (6.9)**
 Visual learninge 45.7 (10.1) 36.0 (19.6) 46.3 (12.1) 32.7 (13.8)**
 Reasoning/problem Solvinge 48.3 (9.7) 28.7 (7.7)** 49.5 (11.6) 40.5 (10.0)**
Serial search behavioral data
 Accuracy 97.6 (8.9) 93.3 (10.9) 100 (0.2) 99 (2.6)
PANSS scoresg
 Positive symptoms – 19.3 (5.4) – 19.7 (5.4)
 Negative symptoms – 16.5 (3.7) – 19.5 (4.2)
 Global psychopathology symptoms – 32.3 (5.6) – 39.4 (8.2)
Open in a new tab

*P < 0.05; **P < 0.01.

Missing data from: a5 patients from PS, 7 patients from NS; b1 patient NS; c1 patient PS, 1 patient and 1 control NS; d1 control NS; e4 controls and 8 patients PS; f3 controls and 1 patient PS; g5 patients NS; and one control in PS for all.

All subjects had corrected 20/32 vision or better on the Logarithmic Visual Acuity Chart (Precision Vision, LaSalle, IL). Symptoms were assessed using Positive and Negative Syndrome Scale (PANSS) (Kay et al. 1987). Neurocognitive function was assessed using the MATRICS consensus cognitive battery (MCCB) (Nuechterlein et al. 2008).

Stimuli

Stimuli were 21° square search-fields with 5.5 cycles/° Gabor-patches, ~1° diameter, displayed against a solid gray background, and randomly distributed within a virtual 8 × 8 grid (e.g., in Fig. 1A). Search-field images were created in Matlab (The Mathworks, Natick, Massachusetts).

In the Behavioral Pilot conjunction-search condition, the target consisted of a low-contrast, vertically oriented Gabor-patch, presented amidst high-contrast (100%), vertical, and low-contrast (3, 5, 7, 10, 20, 40, and 75% of the high contrast) horizontal distractors. Set-sizes varied from 6 to 64 stimuli per search field (6, 12, 24, 48, and 64).

Search fields with all contrasts and set-sizes were presented in pseudorandom order in 8 blocks for a total of 296 trials. Distractor ratio effects were assessed by varying only the number of low-contrast (10% contrast) stimuli; set-sizes (high/low contrast) were as follows: 48 (24/24), 36 (24/12), and 30 (24/6) (Shen et al. 2000). For feature search (parallel), only high-contrast stimuli were used, for 64 trials, presented in two separate blocks.

In the Neurophysiology study (NS) conjunction search, 192 search-fields were presented in blocks of 48 each, with a 5-min rest between blocks. A set-size of 48 stimuli with half at low-(40%) contrast (Shen et al. 2000) was used across search-fields. For feature search, 288 search-fields were presented as the task was faster.

Eye Movements

Search fields were presented on a monitor distant 57 cm from the subject’s eyes (Iiyama Vision Master Pro 512 cathode ray tube (CRT) monitor, refresh rate of 85 Hz, and a screen resolution set to 1024 × 768 pixels). Stimuli presentation was programmed and controlled by Experiment Builder, and the experiment ran either an Eyelink 1000 system, with tower configuration (SR Research Ltd, Canada), or a remote eye-tracking system using monopolar pupil-tracking at 1000 Hz (500 Hz for the NS). Head movement was minimized by use of a chinrest.

Before each block of trials, an instruction screen prompted the participant to look for the lower contrast vertical stimulus. Subjects responded by pressing a button on a Sidewinder Plug & Play Gamepad (Microsoft Corporation). A 9-point calibration procedure and validation procedure was performed at the beginning of the task and was repeated following any breaks.

A drift correction was applied while the subjects fixated on a dot in the center of the screen, followed by a reminder screen indicating the contrast of the upcoming target. The trial ended when the subject fixated on the target and pressed a button on the controller, or timed out after 30 s. To lessen random guessing, the system would only accept a response as correct if a fixation had been made within a 3° window around the target, within 2 s of the button press.

Fixation and saccade measures were obtained offline using DataViewer (SR Research Ltd, Canada). Saccades were defined using a velocity threshold of 30°/s, an acceleration threshold of 8000°/s2, and a minimum motion-threshold of 0.5°. Fixation onset was indexed by saccade end. In the Neurophysiology (but not Behavioral Pilot) study, eye-fixations were also analyzed relative to stimulus locations to determine the characteristic of the stimulus closest to each fixation, as well as the quantitative distance between prefinal and final fixations and target location.

Neurophysiology

Subjects were tested in an electrically shielded, darkened room. Electroencephalography (EEG) data were recorded continuously using an ANT/“Duke” layout Waveguard electrode cap with 64 equidistant scalp electrodes, which increases sampling of occipital regions (Supplementary Fig. S1) and the ANT recording system (ANT Neuro, Einschede, Netherlands) with a sampling rate of 1024 Hz. Impedances were maintained below 10 kΩ. Data were processed offline in MATLAB (Mathworks) using the EEGlab and ERPLAB Toolboxes (Delorme and Makeig 2004; Lopez-Calderon and Luck 2014).

Processing steps included bandpass filtering (0.1–80 Hz), nearest-neighbor bad channel interpolation, and independent component analysis (ICA)-based eye-movement correction. In general, the number of sweeps surviving artifact rejection was higher for patients than controls, reflecting the increased number of fixations needed per search-field for target detection (Supplementary Table S1).

To minimize remaining eye-movement contributions to FRP analyses, data were re-referenced to linked mastoids. Epochs (−200–600 ms) were indexed by fixation onset, and data in any electrode channel (excluding Electrooculography (EOG)) with a change in amplitude greater than 120 μV during a 200 ms time window were excluded from averaging.

Epochs were categorized into “random-search” (i.e., all fixations preceding the prefinal), “prefinal”, and “final” fixations for each search-field, and averaged. Time frequency (TF) measures included evoked amplitude, intertrial coherence (ITC), and single-trial amplitude, and were calculated using 3-cycle and 6-cycle Morlet wavelet convolution for frequencies below and above 25 Hz, respectively.

Before statistical analysis, data were baseline corrected relative to a −150 to −50 ms prefixation interval. An occipitoparietal ROI was defined based upon cross-group topographic maps (Supplementary Fig. S1), and included electrodes POz, Pz, CPz, PO3, PO4, P1, and P2. Predetermined theta (4–8 Hz), delta (0.5–3.5 Hz) and gamma (25–60 Hz) frequency bands were used in the analyses. FRP-related P1 (P1f) and P300 (P300f) amplitudes were measured from 85 to 135 ms and 150 to 300 ms postfixation, respectively. Corresponding TF Integration windows were −50–150 ms (theta) and 150–300 ms (delta). For gamma, separate windows of −50–0 ms (prefixation gamma) and 0–150 (postfixation gamma) were used.

Statistical Analysis

Data were analyzed with SPSS (v24, IBM). RT data were log-transformed. Behavioral data were analyzed by mixed model regression (MMRM) with factors of group and task, and with contrast and density as covariates. Follow-up rmANOVA were performed by task to assess linear effects FRP data were analyzed by rmMANOVA across fixation types (random, prefinal, final).

For across-group correlations, an initial analysis of covariance (ANCOVA) assessed similarity of slopes between the two groups. Across-group partial correlations (rp) or multiple regressions covaried for group were conducted only if the group X covariate ratio in the ANCOVA was nonsignificant. Within group correlations were conducted using Pearson (r) correlations. Gaussian distribution analyses were performed using nonlinear curve fitting in PRISM 5.0 (Graphpad.com). Scan patterns were compared across groups using intraclass correlations for absolute agreement. All statistics are two-tailed with preset α-level for significance of P < 0.05.

Results

Behavioral Pilot Study

An omnibus test across both tasks and all contrasts/set-sizes showed a highly significant main effect of group (F1,36.9 = 16.2, P = 0.0003, d = 1.33). As expected, there were highly significant main effects of contrast (F1,12 300 = 82.2, P < 0.0001) and set-size (F1,12 300 = 135.5, P < 0.0001), as well as significant task X contrast (F1,12 000 = 103.9, P < 0.0001), task X set-size (F1,12 300 = 77.3, P < 0.0001) and task X contrast X set-size (F1,12 000 = 43.9, P < 0.0001) interactions. However, the group X contrast (F1,12 300 = 0.37, P = 0.54), group X set-size (F1,12 300 = 2.67, P = 0.10) and group X contrast X set-size (F1,12 300 = 1.15, P = 0.28) interactions were all nonsignificant.

The group X task interaction was also significant (F1,12 300 = 7.16, P = 0.007), although higher-order interactions involving group were not (all P > 0.2). Separate analyses were, therefore, performed for serial and parallel tasks independently. The main effect of group was independently significant for both the serial (F1,37.9 = 6.72, P = 0.013) and parallel (F1,38.5 = 16.7, P = 0.0002). Group X contrast effects were not significant for either task alone (both P > 0.1). The group X set-size effect was not significant in the serial search condition (F1,8126 = 0.13, P = 0.7) (Fig. 1B), but was significant in the orientation-alone parallel-search condition (F1,4146 = 4.25, P = 0.039). Similar effects were observed in the contrast-alone condition (Supplementary Fig. S2A).

In follow-up analyses, a significant between group difference was observed for fixation number (F1,47 = 4.08, P = 0.049) but not for fixation duration (F1,32.4 = 0.05, P = 0.83) across tasks. These effects were further confirmed in an rmANOVA incorporating mean subject data from the serial search task alone (Supplementary Fig. S2B,C).

Distractor ratio: Across groups, decreasing the ratio of high- to low-contrast distractors from 50 to 20% significantly reduced search time (F2,27 = 52.2, P < 0.0001), with no significant main effect of group (F1,28 = 1.82, P = 0.19) or group X ratio interaction (F1,28 = 0.13, P = 0.88) (Supplementary Fig. S3).

Neurophysiology Study

Serial Search

As in the behavioral PS, patients required a significantly greater amount of time per search-field to detect the targets (F1,47 = 7.39, P = 0.009, d = 0.80) (Fig. 2A), again reflecting a significant increase in the number of fixations per search-field (F1,47 = 7.62, P = 0.008, d = 0.81) (Fig. 2B) with no significant difference in fixation duration (F1,47 = 0.95, P = 0.34) (Fig. 2C).

Figure 2.

Figure 2

Open in a new tab

Behavioral performance during serial search in the NS. (A) Mean (SEM) time per search-field in controls (Ctl) and patients (Pat). (B) Mean number of fixations per search-field. (C) Mean fixation duration. (D) Mean distance between prefinal and final fixation locations and center of the target. The horizontal reference line indicates the border between foveal and parafoveal processing. (E) Scatter plot showing the relationship between Prefinal distance and Search time across groups. The correlation was significant both across groups covaried for group (shown) and in patients alone (P < 0.001). In a separate ANCOVA, the group X covariate interaction was nonsignificant (F1,45 = 2.67, P = 0.11) indicating similar slope across groups. **P < 0.01 patient versus control.

As predicted, prefinal fixation distance—defined as the distance between the last fixation before target detection and the center of the target—was significantly smaller in patients versus controls (F1,47 = 7.84, P = 0.008, d = 0.80) (Fig. 2D), reflecting a smaller visual span. In turn, the increase in prefinal distance strongly predicted mean time per search-field both across groups (Fig. 2E) and in patients alone (r = −.75, P < 0.001). Following covariation for prefinal distance, neither the mean time per search-field (F1,46 = 1.71, P = 0.2) nor the number of fixations (F1,46 = 1.34, P = 0.25) remained significant across groups.

FRP Analyses

FRPs to all fixation types included an initial biphasic wave that preceded fixation onset, followed by a biphasic N1f/P1f component response between 0 and 150 ms. An additional P3f component occurred between 150 and 300 ms for prefinal and final fixations only (Fig. 3A). As opposed to standard visual ERP (Martinez et al. 2018), FRPs both occur earlier and involve phase-reset to a greater degree, suggesting different underlying processes (Supplementary Fig. S4).

Figure 3.

Figure 3

Open in a new tab

Mean occipitoparietal activity to fixation onset for Random-search (top row), prefinal (middle row) and Final (bottom row) fixations. Electrodes used for average: CPz, Pz, POz, P1/P2, and PO3/4. (A) Time–domain potentials showing fixation P1f and P3f potentials, scalp distribution (“headmaps”) and corresponding integration windows. Headmaps for Prefinal fixations were highly similar to those of Random and were omitted. (B–D) TF analyses showing mean evoked amplitude (B), ITC (C) and single-trial analysis amplitude (D) for controls and patients. The black horizontal line indicates transition from 3- to 6-cycle Morlet waveform used in analysis. The red box indicates the integration window for P1f -related theta analyses. The black box indicates the fixation window for P3f -related delta analyses.

In TF analyses (Fig. 3B–D), the prefixation activity corresponded primarily to an increase in gamma activity immediately preceding fixation. By contrast, the N1f/P1f corresponded primarily to an increase in theta activity and the P3f component primarily to an increase in delta (1–4 Hz) activity. An omnibus test across all 3 components and all fixation types showed both a highly significant main effect of group (F1,47 = 11.2, P = 0.002, d = 0.98) and a significant component X group interaction (F2,46 = 4.72, P = 0.014, d = 0.63). Follow-up analyses were, therefore, conducted for each component individually to resolve the interaction.

Prefixation activity/gamma: The amplitude of the prefixation gamma was not significantly different between groups (F1,47 = 0.01, P = 0.92). Both the main effect of fixation type (F2,46 = 2.07, P = 0.14) and the fixation X group interaction (F2,46 = 0.26, P = 0.77) were also nonsignificant (Supplementary Table S2).

P1f/theta: Across all fixation types, both P1 amplitude (F1,47 = 6.67, P = 0.013, d = 0.75) (Supplementary Table S2) and underlying theta activity (F1,47 = 13.3, P = 0.001, d = 1.08) were significantly reduced in patients versus controls across fixation types. The main effect of fixation type (F2,46 = 0.95, P = 0.39) and the fixation-type X group interaction (F2,46 = 0.27, P = 0.77) were nonsignificant (Fig. 4A). For correlational analyses, therefore, theta amplitudes were combined across the three fixation types.

Figure 4.

Figure 4

Open in a new tab

Mean (SEM) FRP-related values from the serial search task as a function of fixation type across groups. (A) Evoked Theta amplitude (corresponding to P1f). (B) Theta ITC. (C) Single-trial theta amplitude. (D) Evoked delta-amplitude (corresponding to P3f). (E) Correlation between evoked theta amplitude and search time in patients. (F) Correlation between evoked theta amplitude and prefinal fixation distance in patients.

In single-trial TF analyses, the between-group difference in theta ITC was highly significant (F1,47 = 20.1, P < 0.0001, d = 1.32) (Fig. 4B). Across groups, the main effect of fixation type was significant with somewhat lower ITC to prefinal or final versus random-search fixations (F1,47 = 6.19, P = 0.016). However, the fixation type X group interaction was nonsignificant (F2,46 = 0.92, P = 0.4).

The between-group difference in single-trial theta amplitude was also significant (F1,47 = 4.32, P = 0.04, d = 0.61) (Fig. 4C), although to a lesser degree. As with ITC, there was a significant main effect of fixation type (F2,46 = 4.06, P = 0.02), with lower single-trial amplitude to random-search fixations than prefinal/final (F1,47 = 8.30, P = 0.006). Between-group differences in theta ITC remained strongly significant even following covariation for amplitude changes (all P < 0.001).

P3f/delta: For both P3f- and evoked delta-amplitude, there was a significant stepwise increase from random-search to prefinal to final fixation. Although the P3f did not significantly discriminate between groups (F1,47 = 1.47, P = 0.23), delta evoked amplitude was significantly reduced in patients vs. controls across all fixation types (F1,47 = 4.73, P = 0.035) (Fig. 4D). The fixation type X group was nonsignificant (F2,46 = 2.21, P = 0.12), suggesting equivalent reductions across the fixation types.

In single-trial analyses, delta ITC also differed significantly across groups across all fixation types (F1,47 = 4.62, P = 0.037) (Supplementary Table S2). Across groups, ITC was not significantly different between random-search and prefinal fixations (F1,47 = 2.40, P = 0.13), but increased significantly from prefinal to final fixation (F1,47 = 94.3, P < 0.0001). The fixation type X group interaction was nonsignificant (F2,46 = 1.31, P = 0.28), reflecting the parallel increases between groups.

By contrast, delta single-trial amplitude did not differ significantly between groups (F1,47 = 2.47, P = 0.12). Across groups, a progressive increase was observed from random-search to prefinal (F1,47 = 36.6, df = 1,47, P < 0.0001) and prefinal to final (F1,47 = 60.8, P < 0.0001) fixation. The fixation type X group interaction (F2,46 = 0.28, P = 0.75) was nonsignificant.

Correlation between measures: Across groups (covaried for group membership), evoked theta amplitude correlated significantly with both total search time (rp = −.34, P = 0.018) and prefinal fixation distance (rp = 0.39, P = 0.008). Significant correlations were also observed with the schizophrenia group alone for both total search time (Fig. 4E) and prefinal fixation distance (Fig. 4F). By contrast, no significant correlations were observed with the healthy controls (HC) group alone (both P > 0.1).

Theta ITC also correlated significantly with both total search time (rp = −.31, P = 0.033) and prefinal fixation distance (rp = 0.38, P = 0.007), while correlations with single-trial amplitude were nonsignificant. The correlation between ITC and prefinal fixation distance was also highly significant within the patient group alone (r = 0.53, P = 0.008).

Across groups, reductions in evoked delta-amplitude correlated strongly with reductions in evoked theta (rp = 0.50, P = 0.004). Correlations were significant within the patient (r = 0.46, P = 0.023) and control (r = 0.52, P = 0.007) groups, respectively. Following covariation for theta, deficits in delta were no longer significant between groups (F1,46 = 0.12, P = 0.73) (Supplementary Fig. S5).

NS, Parallel Search

As in the serial search task, patients showed a significant increased mean search time (F1,47 = 5.60, P = 0.022) as well as number of fixations (F1,47 = 5.38, P = 0.025). Mean fixation duration was not significantly different (F1,47 = 3.15, P = 0.083) (Supplementary Table S4).

Also, as in the serial search task, the distance from prefinal fixation to the target was significantly smaller in patients vs. controls (F1,47 = 7.67, P = 0.008), and between-group differences in search time (F1,47 = 1.02, P = 0.32) and number of fixations per search-field (F1,47 = 0.89, =0.35) between groups were no longer significant following covariation for prefinal fixation distance.

In a rmANOVA across the serial and parallel tasks, the main effect of group for prefinal fixation distance was highly significant (F1,47 = 8.35, P = 0.006) as was the main effect of task (F1,47 = 238.9, P < 0.0001). By contrast, the group X task interaction was nonsignificant (F1,47 = 2.43, P = 0.13).

FRP: In the parallel-search task, because of the limited number of saccades/fixations per search-field, FRPs were only resolved to prefinal and target fixations (Supplementary Fig. S6). As in the serial search task, a prominent P1f component was observed, and was reduced in schizophrenia (F1,47 = 4.03, P = 0.05). In TF analyses, reductions were observed in both evoked theta amplitude (F1,47 = 4.15, P = 0.047) and ITC (F1,47 = 5.72, P = 0.021). (Supplementary Table S4). P1f amplitudes correlated significantly with search time (rp = −.55, P < 0.0001), number of fixations (rp = −0.37, P = 0.01) and penultimate fixation distance (rp = 0.54, P < 0.0001) across groups (covaried for group), as well as in patients individually.

Across-Task Comparisons

A final analysis evaluated the statistical distribution of prefinal fixation distances for serial versus parallel-search tasks across groups. For all subjects and both tasks, mean prefinal fixation distances fell within the parafoveal/perifoveal span (1–8°). All distributions were accounted for well by a single Gaussian distribution (all R2 > 0.6) (Fig. 5A,B). Mean values were significantly larger for parallel than serial across both groups. For both serial (F1,18 = 40.0, P < 0.0001) and parallel (F1,18 = 9.59, P = 0.006) tasks the mean distance was shifted significantly to the left in patients, suggesting reduced visual span.

Figure 5.

Figure 5

Open in a new tab

Fixation analyses. (A,B) Distributions of prefinal fixation distances to the target for the serial and parallel-search tasks for controls (A) and patients (B). (C) Correlation between prefinal distances on the serial and parallel-search tasks, across groups. (D) Correlation between patients vs. controls fixation time as a function of grid location. (E) Relative number of fixations to low- and high-contrast distractors in patients and controls. ***P < 0.000. (F) Percentage of fixations on low-contrast stimuli for both groups, showing no significant difference between groups.

In a 2-way analysis of variance across tasks, the main effects of group (F1,47 = 9.01, P = 0.004) and task (F1,47 = 272.8, P < 0.0001) were strongly significant, whereas the interaction effect was not (F1,47 = 0.27, P = 0.61). Moreover, strong correlations were observed across subjects for the two measures (Fig. 5C).

In both the serial and parallel-search tasks, patients showed similar scanning patterns to those of controls across search-fields with highly significant correlations between groups (Fig. 5D, Supplementary Fig. S7). Furthermore, both controls and patients equivalently “oversampled” the low-contrast stimuli during serial search, as shown by a significantly increased number of fixations near low- versus high-contrast stimuli across groups (F1,47 = 22.7, P < 0.0001) (Fig. 5E). Although the total number of fixations per search-field was significantly higher in patients than controls (F1,47 = 7.50, P = 0.009), the stimulus-type X group interaction was nonsignificant (F1,47 = 1.30, P = 0.26). When calculated as percentages, both groups showed a similar preferential sampling of the low-contrast stimuli, with no significant difference (t = −.06, P = 0.95) between-groups (Fig. 5F), suggesting similar search strategies.

Clinical Correlations

As expected, patients showed a highly significant reduction in neurocognitive function as assessed using the MCCB (F1,46 = 39.7, P < 0.0001, d = 1.8) (Table 1). Within patients, evoked theta response correlated with attention-vigilance (r = 0.56, P = 0.005) and speed of processing (SoP, r = 0.50, P = 0.014) (Supplementary Fig. S8). By contrast, correlations with nonvisual components of the MCCB, such as auditory working memory or reasoning/problem solving, were nonsignificant (P > 0.4). The correlation with attention-vigilance, but not SoP, remained significant even following control for multiple comparisons (n = 6, critical P = 0.008).

Delta-amplitude correlated significantly with negative symptoms (P < 0.028). However, the correlation did not remain significant following control for multiple comparisons. All other correlations to either symptoms or medication dosage were nonsignificant.

Discussion

Deficits in auditory and visual sensory processing in schizophrenia have become increasingly appreciated over recent years (Javitt 2009; Javitt and Freedman 2015). Nevertheless, the degree to which these affect naturalistic functions remains to be determined. Here, we used a visual Guided-Search task combined with eye-tracking and EEG to investigate contributions of visual active sensing to behavioral disturbance. Although visual Guided Search has been endorsed as an informative paradigm for investigating neural mechanisms underlying attentional allocation and sensory processing impairments in schizophrenia (Nuechterlein et al. 2009), to our knowledge, this is the first study to utilize combined eye-tracking and neurophysiological measures to investigate underlying neural mechanisms.

As expected, patients showed substantially increased search times compared to controls in both the serial- and parallel-search versions of the task (VanMeerten et al. 2016). In both cases, the increased search time was associated with an increase in the number—but not duration—of fixations per search-field, which in turn was associated with a markedly (P = 0.008) reduced distance, over which targets could be detected in peripheral (parafoveal) vision relative to the point of fixation (Elahipanah et al. 2011). Once the decrease in size of the visual “spotlight” (visual span) during search was considered, increased search times on this task were no longer significant.

At the neurophysiological level, the reduced visual span was associated with reduced FRP generation over posterior visual cortex (Fig. 3). Consistent with primate studies, the human FRP response corresponded specifically to a phase-reset of ongoing theta-rhythms with little alteration in single-trial power (Rajkai et al. 2008). Overall, these findings are consistent with a model in which impaired visual active sensing—as reflected in reduced FRP—leads to narrowing of the visual span which, in turn, drives behavioral impairment (Fig. 4E,F). These findings also converge with a growing body of literature suggesting theta response abnormalities in schizophrenia across both auditory (Lee et al. 2017; Javitt et al. 2018) and visual (Martinez et al. 2015) sensory systems, and support prior studies demonstrating reduced stimulus-related processing during visual search versus nonsearch tasks in schizophrenia (Davenport et al. 2006; VanMeerten et al. 2016).

By contrast to the deficits in FRP and visual span, the slope of the RT increase as a function of set-size did not differ significantly between groups (Fig. 1B). Furthermore, patients and controls fixated on similar areas of search-fields (Supplementary Fig. S7), and both groups equivalently “oversampled” the low- versus high-contrast distractors (Fig. 5E). In addition, both groups showed similar effects of distractor ratio manipulation (Supplementary Fig. S3), consistent with a prior report (Elahipanah et al. 2008).

Prior studies of visual search in schizophrenia have yielded mixed results in the “target present” condition such as used here (Mori et al. 1996; Fuller et al. 2006; Gold et al. 2007; Tanaka et al. 2007), potentially related to methodological issues. For example, in prior studies between-group differences may also have been driven in part by use of raw, rather than log-transformed RT data, which are inherently right-skewed. Also, different stimuli were used across studies. Here, stimulus contrast was detectable from each Gabor-patch as a whole, encouraging its use to preselect stimuli for further evaluation, whereas orientation could only be resolved by more detailed examination.

However, it remains to be determined whether the use of other stimulus features to guide processing (e.g., color, shape) may have resulted in changes in slope as well as impairments related primarily to visual-level processing, reflecting additional “top–down” impairments. The absence of physiological data in prior studies also limits comparability across studies.

Unexpectedly, in secondary analyses of the parallel-search data from the behavioral pilot we observed a significant interaction between group and set-size, suggesting potential different strategies for performing the task. The finding should be treated cautiously as the group X set-size interaction was not significant in the primary analyses or in either group independently. The finding may nevertheless deserve further investigation.

FRP: The P1f (λ potential) has been the focus of increasing investigation in a range of contexts including Guided-Search tasks and naturalistic reading in normal volunteers (Kaunitz et al. 2014). However, to our knowledge, this is the first study of FRP generation in schizophrenia and also the first study to use TF analyses of the FRP in humans to analyze underlying mechanisms. The P1f potential shows a similar scalp distribution to that of the stimulus-driven P1 potential, which we (Doniger et al. 2002; Friedman et al. 2012; Javitt 2015) and others (Luck et al. 2006) have also found to be reduced in schizophrenia. Both components also map similarly into the theta-frequency range. However, the intracortical mechanisms by which they are generated differ significantly (Rajkai et al. 2008).

Specifically, although both the P1 and P1f originate from visual sensory cortex, the laminar profile of activity differs such that the stimulus-evoked P1 is driven primarily by thalamocortical input into layer 4, whereas the P1f is driven disproportionately by nonthalamic inputs that bypass layer 4 (Rajkai et al. 2008). At present, the source of these inputs is unknown, and could include top–down projections from frontoparietal cortex, or bottom–up projections from subcortical (e.g., lateral–pulvinar, intranuclear–thalamic) or brainstem regions (Rajkai et al. 2008).

In both intracranial recordings (Rajkai et al. 2008) and our study, the P1f consists primarily of fixation-driven phase-reset of the ongoing theta-rhythms in visual cortex, in contrast to the stimulus-evoked P1, which shows proportionate theta-band increases in both ITC and power (e.g., Rajkai et al. 2008; Martinez et al. 2018; Supplementary Fig. S4). Nevertheless, future studies investigating the relationship between impaired P1 and P1f generation in schizophrenia are needed.

In addition to the reduction in P1f, deficits were observed in the P3f, which occurs only to prefinal, and, especially, to final fixations and is thus likely related to target detection. During search, amplitudes of both P1f and P3f can be enhanced by increasing attentional demands, for example by introducing a competing auditory task (Ries et al. 2016). In such conditions, however, the amplitudes of the P1f and P3f components increase in parallel, along with P3f latency. Here, P1f was diminished disproportionately to P3f with no significant between group difference in accuracy and no prolongation of P3f latency.

Present findings converge with recent suggestions that theta-phase significantly modulates local gain within sensory regions (Fiebelkorn and Kastner 2019) and contributes to integration versus segregation of visual information (Wutz et al. 2016), as well as with recent findings of impaired auditory “active sensing” in schizophrenia (Lakatos et al. 2013). As opposed to gamma rhythms that interrelate primarily with parvalbumin-type gamma-aminobutyric acid interneurons in cortex, theta-rhythms interrelate additionally with somatostatin-type interneurons. Optogenetic silencing of somatostatin-, but not parvalbumin-, interneurons in rodent visual cortex reproduce patterns of visual ERP deficits observed in schizophrenia (Hamm and Yuste 2016).

Limitations

Sample sizes for the present study are relatively limited. Although significant between-group effects were observed both in the behavioral pilot (n = 15/group) and in the neurophysiological study (n = 25/group), the degree to which results will replicate within larger samples remains to be determined. Also, all subjects were on medication at the time of testing. Although no correlations were observed between our experimental measures and antipsychotic dose measured in chlorpromazine (CPZ) equivalents, an effect of medication cannot be excluded. Finally, although we postulate that deficits in FRP generation in Sz reflect impaired subcortical input into primary visual cortex, future neuroimaging studies are required to interrogate underlying circuits.

Overall, the present study provides the first demonstration of impaired visual active sensing in schizophrenia, echoing recent findings in the auditory system (Lakatos et al. 2013). Impaired sensory processing represents a critical, yet understudied, contributor to impaired neurocognitive dysfunction in schizophrenia. Correlations between reduced FRP amplitude and visually dependent cognitive domains, such as SoP and attention-vigilance highlight the importance of visual active sensing deficits to overall patterns of neurocognitive impairments in schizophrenia.

Supplementary Material

Supplementary material is available at Cerebral Cortex online.

Funding

The National Institutes of Health (MH049334 to D.C.J.); National Research Council of Brazil (CNPq and CAPES to Dr F.B.).

Notes

We thank Gail Silipo, Walter Machado-Pinheiro, and José Magalhães de Oliveira for their assistance.

Supplementary Material

Supplementary_Data_bhz277
Click here for additional data file. (1.2MB, pdf)

References

  1. Bedwell JS, Chan CC, Cohen O, Karbi Y, Shamir E, Rassovsky Y. 2013. The magnocellular visual pathway and facial emotion misattribution errors in schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry. 44:88–93. [DOI] [PubMed] [Google Scholar]
  2. Butler PD, Abeles IY, Weiskopf NG, Tambini A, Jalbrzikowski M, Legatt ME, Zemon V, Loughead J, Gur RC, Javitt DC. 2009. Sensory contributions to impaired emotion processing in schizophrenia. Schizophr Bull. 35:1095–1107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Concha-Miranda M, Rios J, Bou J, Valdes JL, Maldonado PE. 2019. Timing is of the essence: improvement in perception during active sensing. Front Behav Neurosci. 13:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Davenport ND, Sponheim SR, Stanwyck JJ. 2006. Neural anomalies during visual search in schizophrenia patients and unaffected siblings of schizophrenia patients. Schizophr Res. 82:15–26. [DOI] [PubMed] [Google Scholar]
  5. Delorme A, Makeig S. 2004. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 134:9–21. [DOI] [PubMed] [Google Scholar]
  6. Dias EC, Butler PD, Hoptman MJ, Javitt DC. 2011. Early sensory contributions to contextual encoding deficits in schizophrenia. Arch Gen Psychiatry. 68:654–664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Doniger GM, Foxe JJ, Murray MM, Higgins BA, Javitt DC. 2002. Impaired visual object recognition and dorsal/ventral stream interaction in schizophrenia. Arch Gen Psychiatry. 59:1011–1020. [DOI] [PubMed] [Google Scholar]
  8. Elahipanah A, Christensen BK, Reingold EM. 2008. Visual selective attention among persons with schizophrenia: the distractor ratio effect. Schizophr Res. 105:61–67. [DOI] [PubMed] [Google Scholar]
  9. Elahipanah A, Christensen BK, Reingold EM. 2011. Controlling the spotlight of attention: visual span size and flexibility in schizophrenia. Neuropsychologia. 49:3370–3376. [DOI] [PubMed] [Google Scholar]
  10. Fiebelkorn IC, Kastner S. 2019. A rhythmic theory of attention. Trends Cog Sci; 23:87–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fourment A, Calvet J, Bancaud J. 1976. Electrocorticography of waves associated with eye movements in man during wakefulness. Electroencephalogr Clin Neurophysiol. 40:457–469. [DOI] [PubMed] [Google Scholar]
  12. Friedman T, Sehatpour P, Dias E, Perrin M, Javitt DC. 2012. Differential relationships of mismatch negativity and visual p1 deficits to premorbid characteristics and functional outcome in schizophrenia. Biol Psychiatry. 71:521–529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fuller RL, Luck SJ, Braun EL, Robinson BM, McMahon RP, Gold JM. 2006. Impaired control of visual attention in schizophrenia. J Abnorm Psychol. 115:266–275. [DOI] [PubMed] [Google Scholar]
  14. Gold JM, Fuller RL, Robinson BM, Braun EL, Luck SJ. 2007. Impaired top-down control of visual search in schizophrenia. Schizophr Res. 94:148–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gracitelli CP, Vaz de Lima FB, Bressan RA, Paranhos JA. 2013. Visual field loss in schizophrenia: evaluation of magnocellular pathway dysfunction in schizophrenic patients and their parents. Clin Ophthalmol. 7:1015–1021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hamm JP, Yuste R. 2016. Somatostatin interneurons control a key component of mismatch negativity in mouse visual cortex. Cell Rep. 16:597–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hessels RS, Niehorster DC, Nystrom M, Andersson R, Hooge ITC. 2018. Is the eye-movement field confused about fixations and saccades? A survey among 124 researchers. R Soc Open Sci. 5:180502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ito J, Maldonado P, Grun S. 2013. Cross-frequency interaction of the eye-movement related LFP signals in V1 of freely viewing monkeys. Front Sys Neurosci. 7:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Javitt DC. 2009. When doors of perception close: bottom-up models of disrupted cognition in schizophrenia. Annu Rev Clin Psychol. 5:249–275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Javitt DC. 2015. Neurophysiological models for new treatment development in schizophrenia: early sensory approaches. Ann N Y Acad Sci. 1344:92–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Javitt DC, Freedman R. 2015. Sensory processing dysfunction in the personal experience and neuronal machinery of schizophrenia. Am J Psychiatry. 172:17–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Javitt DC, Lee M, Kantrowitz JT, Martinez A. 2018. Mismatch negativity as a biomarker of theta band oscillatory dysfunction in schizophrenia. Schizophr Res. 191:51–60. [DOI] [PubMed] [Google Scholar]
  23. Kamienkowski JE, Ison MJ, Quiroga RQ, Sigman M. 2012. Fixation-related potentials in visual search: a combined EEG and eye tracking study. J Vis. 12:4. [DOI] [PubMed] [Google Scholar]
  24. Kaunitz LN, Kamienkowski JE, Varatharajah A, Sigman M, Quiroga RQ, Ison MJ. 2014. Looking for a face in the crowd: fixation-related potentials in an eye-movement visual search task. Neuroimage. 89:297–305. [DOI] [PubMed] [Google Scholar]
  25. Kay SR, Fiszbein A, Opler LA. 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull. 13:261–76. [DOI] [PubMed] [Google Scholar]
  26. Kazai K, Yagi A. 2003. Comparison between the lambda response of eye-fixation-related potentials and the P100 component of pattern-reversal visual evoked potentials. Cogn Affect Behav Neurosci. 3:46–56. [DOI] [PubMed] [Google Scholar]
  27. Kim DW, Shim M, Song MJ, Im CH, Lee SH. 2015. Early visual processing deficits in patients with schizophrenia during spatial frequency-dependent facial affect processing. Schizophr Res. 161:314–321. [DOI] [PubMed] [Google Scholar]
  28. Konig P, Luksch H. 1998. Active sensing-closing multiple loops. Z Naturforsch C. 53:542–549. [DOI] [PubMed] [Google Scholar]
  29. Lakatos P, Schroeder CE, Leitman DI, Javitt DC. 2013. Predictive suppression of cortical excitability and its deficit in schizophrenia. J Neurosci. 33:11692–11702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lee M, Sehatpour P, Hoptman MJ, Lakatos P, Dias EC, Kantrowitz JT, Martinez AM, Javitt DC. 2017. Neural mechanisms of mismatch negativity dysfunction in schizophrenia. Mol Psychiatry. 22:1585–1593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Leonard CJ, Robinson BM, Hahn B, Gold JM, Luck SJ. 2014. Enhanced distraction by magnocellular salience signals in schizophrenia. Neuropsychologia. 56:359–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lopez-Calderon J, Luck SJ. 2014. ERPLAB: an open-source toolbox for the analysis of event-related potentials. Front Hum Neurosci. 8:213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Luck S, Fuller R, Braun E, Robinson B, Summerfelt A, Gold J. 2006. The speed of visual attention in schizophrenia: electrophysiological and behavioral evidence. Schizophr Res. 85:174–195. [DOI] [PubMed] [Google Scholar]
  34. Martinez A, Gaspar PA, Hillyard SA, Andersen SK, Lopez-Calderon J, Corcoran CM, Javitt DC. 2018. Impaired motion processing in schizophrenia and the attenuated psychosis syndrome: etiological and clinical implications. Am J Psychiatry. 175:1243–1254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Martinez A, Gaspar PA, Hillyard SA, Bickel S, Lakatos P, Dias EC, Javitt DC. 2015. Neural oscillatory deficits in schizophrenia predict behavioral and neurocognitive impairments. Front Hum Neurosci. 9:371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Martinez A, Hillyard SA, Bickel S, Dias EC, Butler PD, Javitt DC. 2012. Consequences of magnocellular dysfunction on processing attended information in schizophrenia. Cereb Cortex. 22:1282–1293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mori S, Tanaka G, Ayaka Y, Michitsuji S, Niwa H, Uemura M, Ohta Y. 1996. Preattentive and focal attentional processes in schizophrenia: a visual search study. Schizophr Res. 22:69–76. [DOI] [PubMed] [Google Scholar]
  38. Nobre AC, Coull JT, Walsh V, Frith CD. 2003. Brain activations during visual search: contributions of search efficiency versus feature binding. Neuroimage. 18:91–103. [DOI] [PubMed] [Google Scholar]
  39. Nuechterlein KH, Green MF, Kern RS, Baade LE, Barch DM, Cohen JD, Essock S, Fenton WS, Frese FJ 3rd, Gold JM et al. 2008. The MATRICS consensus cognitive battery, part 1: test selection, reliability, and validity. Am J Psychiatry. 165:203–213. [DOI] [PubMed] [Google Scholar]
  40. Nuechterlein KH, Luck SJ, Lustig C, Sarter M. 2009. CNTRICS final task selection: control of attention. Schizophr Bull. 35:182–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Poletti M, Rucci M, Carrasco M. 2017. Selective attention within the foveola. Nat Neurosci. 20:1413–1417. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rajkai C, Lakatos P, Chen CM, Pincze Z, Karmos G, Schroeder CE. 2008. Transient cortical excitation at the onset of visual fixation. Cereb Cortex. 18:200–209. [DOI] [PubMed] [Google Scholar]
  43. Ries AJ, Touryan J, Ahrens B, Connolly P. 2016. The impact of task demands on fixation-related brain potentials during guided search. PLoS One. 11:e0157260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Schroeder CE, Wilson DA, Radman T, Scharfman H, Lakatos P. 2010. Dynamics of active sensing and perceptual selection. Curr Opin Neurobiol. 20:172–176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Shen J, Reingold EM, Pomplun M. 2000. Distractor ratio influences patterns of eye movements during visual search. Perception. 29:241–250. [DOI] [PubMed] [Google Scholar]
  46. Tanaka G, Mori S, Inadomi H, Hamada Y, Ohta Y, Ozawa H. 2007. Clear distinction between preattentive and attentive process in schizophrenia by visual search performance. Psychiatry Res. 149:25–31. [DOI] [PubMed] [Google Scholar]
  47. Treisman A. 1982. Perceptual grouping and attention in visual search for features and for objects. J Exp Psychol Hum Percept Perform. 8:194–214. [DOI] [PubMed] [Google Scholar]
  48. VanMeerten NJ, Dubke RE, Stanwyck JJ, Kang SS, Sponheim SR. 2016. Abnormal early brain responses during visual search are evident in schizophrenia but not bipolar affective disorder. Schizophr Res. 170:102–108. [DOI] [PubMed] [Google Scholar]
  49. Wutz A, Muschter E, Koningsbruggen MG, Weisz N, Melcher D. 2016. Temporal integration windows in neural processing and perception aligned to saccadic eye movements. Curr Biol: CB. 26:1659–1668. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary_Data_bhz277
Click here for additional data file. (1.2MB, pdf)

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press

RESOURCES