Widespread hemodynamic responses, time-locked to trial onsets, occur in the absence of a visual stimulus in early visual cortex, the earliest site of visual cortical processing (Sirotin and Das, 2009). These ‘task-related’ responses (TRRs) have been reported in awake macaques (Sirotin and Das, 2009; Cardoso et al., 2012; Cardoso et al., 2019) using intrinsic signal optical imaging, and in humans using BOLD-fMRI (Ress et al., 2000; Jack et al., 2006; Roth et al., 2020). Unlike stimulus-evoked responses in V1, they are poorly predicted by changes in mean firing rates or local field potential (LFP) amplitudes (Sirotin and Das, 2009), and unlike spatial attentional responses (Roth et al., 2020; Tootell et al., 1998; Gandhi et al., 1999; Herrmann et al., 2010), they are spatially diffuse, extending far beyond the focus of spatial attention (Sirotin and Das, 2009; Roth et al., 2020). Previous studies demonstrate that TRRs are modulated in amplitude by reward and correlate with heart rate and pupil size (Cardoso et al., 2019; Roth et al., 2020), suggesting that they reflect arousal. These findings raise the question of whether TRRs in early visual cortex change from trial to trial according to task difficulty and behavioral performance, variables linked to arousal level (Burlingham et al., 2022).

In this study, we measured TRRs in human early visual cortex while observers performed a visual orientation discrimination task varying in difficulty, building on the task protocol introduced by Roth et al., 2020. Roth et al., 2020 averaged fMRI responses over many trials and observed a positive relationship between reward magnitude and trial-averaged TRR amplitude. Here, we instead used a general linear mixed model (GLMM) to probe the relation between behavioral performance and response amplitude on the timescale of individual trials (Chen et al., 2013). We found that TRRs in early visual cortex scaled in amplitude with task difficulty, accuracy, reaction time (RT), and lapses, with decreasing strength ascending the visual cortical hierarchy. We found similar modulations in pupil size and cardiac activity. Importantly, task- and behavior-dependent modulations of the TRR were not explained by the artifactual influence of respiration and cardiac activity on cerebral blood oxygenation, which suggests that TRRs are linked directly to arousal. Our results demonstrate the existence of a widespread hemodynamic response in early visual cortex that reflects arousal on the timescale of individual trials. One intriguing possibility raised by our findings is that this arousal-linked response may alter encoding and/or decoding of visual information, thereby having a causal influence on behavior. For example, this might occur by modulating the cortical circuit computation to shift the balance in inference between prior and likelihood (Dayan and Yu, 2006; Heeger, 2017; de Gee et al., 2020).

We observed large fMRI responses in human early visual cortex that were spatially widespread, independent of visual stimulation, and tightly entrained to trial timing. These TRRs, measured in V1 in the hemisphere ipsilateral to the stimulus, were larger in amplitude when the task was more difficult, and on trials on which observers made an erroneous response, failed to respond, or were faster to respond. Neither the TRR nor its modulation with task difficulty or behavioral performance was explained by common fMRI-BOLD artifacts, i.e., changes in head movement, cardiac activity, and respiration, nor by the fMRI ‘global signal’. The temporal variability of the TRR was also modulated by task difficulty and behavioral performance, complementing similar findings from previous studies that manipulated reward (Cardoso et al., 2019; Roth et al., 2020). Contrary to the consensus of opinion that the fMRI-BOLD signal in V1 primarily reflects stimulus-evoked and attentional neuronal activity, we found evidence that the BOLD signal also contains a component that reflects the demands of a task and behavioral performance on the timescale of individual trials.

TRRs in early visual cortex are modulated by task difficulty, reward, and behavioral performance. Two previous studies Cardoso et al., 2019 and Roth et al., 2020 examined the effect of reward magnitude on TRRs in early visual cortex, and found that larger rewards led to decreased temporal variability and increased response amplitude on individual trials, as well as larger trial-averaged response amplitude. Roth et al., 2020 found that multiple measures of trial-to-trial variability in TRRs (i.e. timepoint, amplitude, and temporal variability) changed with reward. The authors used computational simulations to show that simultaneous increases in trial-averaged TRR amplitude with reward and decreases in the three measures of variability were consistent with only one possibility: changes in trial-to-trial temporal jitter. That is, averaging over many temporally jittered responses led to a lower amplitude trial-averaged response whereas averaging over many temporally aligned responses led to the opposite. And likewise, more temporal jitter also generated increases in all measures of trial-to-trial variability (timepoint, amplitude, and temporal variability). However, these results don’t preclude additional changes in trial-to-trial response amplitude with reward, but instead suggest that a portion of the trial-averaged response amplitude cannot be predicted by trial-to-trial amplitude alone. Indeed, Cardoso et al. found that both the trial-to-trial amplitude and timing of TRRs in macaque V1 were modulated by reward (Cardoso et al., 2019). Our results provide a complementary account. We found that trial-to-trial amplitude, trial-averaged amplitude, and temporal variability of the TRR were all modulated by task difficulty and behavioral performance. Temporal variability was reduced for unexpected task events (i.e. incorrect feedback tone on an easy run of trials), consistent with previous findings showing that temporal precision (1/variability) scaled with arousal (Cardoso et al., 2019; Roth et al., 2020). Taken together, the evidence suggests that reward, task difficulty, and behavioral performance modulate both the temporal variability and amplitude of TRRs.

We found that TRR amplitude was modulated by RT on a trial-to-trial basis. It is well known that arousal influences RTs, with a classic inverted-U relation (Widdicombe and Sterling, 1970), an effect that shows up in standard physiological measures of arousal, including pupil size (Burlingham et al., 2022) and cardiac activity (Chin and Kales, 2019). So if TRR amplitude reflects arousal, as we hypothesized, we would expect that it would be higher when an observer responds faster, which is what we found (on correct trials). Note that we assume that our task primarily operates in the left half of the inverted-U, because even the hard task is rather boring, so arousal levels are never very high (Burlingham et al., 2022). Previous fMRI studies offer an alternative, but complementary perspective on the relationship between RT and TRR amplitude. A previous study measured the TRR in human V1 in a visual detection task (Jack et al., 2006) with either an immediate or delayed behavioral report (delay, ~9 s). In the delayed condition, there were response peaks following both trial onset and the behavioral report. In peripheral V1, far from the representation of the stimulus, the amplitude of fMRI activity was lower for a delayed than immediate report. This indicates that for an immediate report, there were also two components of the TRR — one for the trial onset and one for the response — but because they were so close in time, their impulse responses summed to a unimodal response. We might assume that two such ‘blurred-together’ responses were also present in our experiment, in which observers responded immediately after the stimulus presentation. It would follow then, that differences in RT should affect this summation and thereby modulate the amplitude of the unimodal TRR — slower RTs associated with lower fMRI amplitudes (Figure 3—figure supplement 2A, Figure 3—figure supplement 3B). Conversely, it could be that the amplitude of the TRR reflects time-on-task, i.e., a sustained input between trial onset and RT, often modeled as a boxcar (Yarkoni et al., 2009). If true, TRR amplitude would be higher for trials with slower RTs (Figure 3—figure supplement 2B, Figure 3—figure supplement 3C). And a third possibility is that all three inputs are present, at trial onset, at the response time, and time on task. This scenario would produce some effect in-between these two extremes, which depends flexibly on the relative strengths of the three inputs. A model similar to this, with three inputs, is commonly used to predict the task-evoked pupil response from task structure (Burlingham et al., 2022; de Gee et al., 2014; Denison et al., 2020). To be liberal in our assumptions, we used a GLMM with these three putative inputs — fMRI_TO, fMRI_BP, and fMRI_ToT — as well as their combinations. We found that the combination of all three inputs best explained the results (i.e. the interactions of the combined fMRI predictor with task and behavioral variables were significant in the model). This agrees with the results of many studies linking arousal and pupil size, which used a similar model (de Gee et al., 2017; de Gee et al., 2014; Denison et al., 2020; van den Brink et al., 2016). Differences in RT are associated with changes in the timing and variability of phasic locus coeruleus (LC) activity (Aston-Jones and Cohen, 2005). And pupil size changes are linked closely to changes in neural activity in LC (Joshi et al., 2016) as well as other neuromodulatory brainstem loci, including the dorsal raphé (Cazettes et al., 2021), ventral tegmental area, subsantia nigra, and basal forebrain, and other subcortical and cortical structures (Burlingham et al., 2022; de Gee et al., 2017; Joshi et al., 2016). Thus, our results suggest that the link between TRR amplitude and RT is mediated by neuromodulatory influences on visual cortex and/or other, decision-related areas of the brain.

We found that TRRs and pupil responses were both modulated by task difficulty and behavioral accuracy in a similar manner. Previous empirical and computational modeling work suggests that such modulations may emerge via pupil-linked arousal systems being driven by internal beliefs about ones own behavioral choices (Urai et al., 2017; Colizoli et al., 2018). Signal detection theoretic models that take into account ones uncertainty about ones own choices (i.e. confidence) correctly predict modulations of the task-evoked pupil response before and after external feedback and their scaling with difficulty and accuracy (Urai et al., 2017; Colizoli et al., 2018). According to these models, the brain might use its arousal system to ‘broadcast’ these uncertainty signals to a wide set of brain regions in order to facilitate inference and information processing. Under this account, the TRR’s modulation with difficulty and behavior may reflect adaptations in an underlying computational process to task demands, internal sensory noise, and internal beliefs (decision uncertainty). Accordingly, we found modulations of the task-evoked pupil response and TRR with difficulty and accuracy that matched predictions from this model (compare our Figures 4B and 5B with Figure 2B from Colizoli et al., 2018 ). This decision-uncertainty-arousal model also predicts that RTs vary in the same manner (i.e. fastest to slowest: easy correct, hard correct, hard incorrect, easy incorrect) — however, our findings weren’t consistent with this. Only 1/9 observers showed the expected pattern of RTs; instead, RTs were faster overall for easy trials, regardless of accuracy (Figure 3D, inset). This difference may be due to the fact that there was no variability in stimulus orientation in our easy task, but there was in the hard task (due to the staircase), whereas the stimulus variability was the same across difficulty levels in refs (Urai et al., 2017; Colizoli et al., 2018). A ripe area for future study would be to link decision-theoretic models such as this with a linear systems model of the fMRI signal, and try to predict the TRR from task and behavioral parameters.

Our results are not attributable to the effects of spatial attention. Previous studies have measured the spatial distribution of spatial attention (the ‘attention field’ ) in early visual cortex and found that it tightly surrounds the retinotopic location of the attended stimulus (Herrmann et al., 2010; Carandini and Heeger, 2011), particularly when it appears at a predictable location, size, and time, as in our experiment. Furthermore, we previously found, using a similar stimulus and protocol, that reward magnitude modulated TRR amplitude (also measured in ipsilateral cortex) even though behavioral accuracy and RT were statistically indistinguishable between low and high reward conditions (Roth et al., 2020). That is, increased reward did not alter the allocation of spatial attention, as quantified as a behavioral enhancement, which is how attention is typically operationalized (Dugué et al., 2020). In our study, we found that the measured TRR was lower in amplitude ascending the visual cortical hierarchy. This is opposite to what is seen in fMRI studies of spatial attention responses, which increase in amplitude from V1 to V2 to V3 (Dugué et al., 2020).

The origin and control of the TRR in early visual cortex remain unclear. One possibility is that they are driven by an intrinsic vascular mechanism rather than a local neuronal one (Sirotin and Das, 2009; Cardoso et al., 2012; Cardoso et al., 2019; Bekar et al., 2012; Sirotin et al., 2012; Herman et al., 2017; Das et al., 2021). This putative mechanism, vasomotion, or oscillation in tone of blood vessels independent of cardiac and respiratory activity, occurs in cerebral arteries and is influenced by arousal (Nilsson and Aalkjaer, 2003; Borovik et al., 2005; Mateo et al., 2017). More broadly, cerebral blood flow is complex and modulated by arousal via multiple cellular mechanisms (Bekar et al., 2012; Raichle et al., 1975), including activity of astrocytes (Duffy and MacVicar, 1995; Bekar et al., 2008; Takata et al., 2015; Wang et al., 2018), pericytes (Peppiatt et al., 2006; Hall et al., 2014), endothelial cells (Sorriento et al., 2011), and smooth muscle cells (Faber, 1988). Our fMRI data cannot reveal the origins of the TRR, but may provide some clues. We found that the amplitude of the trial-averaged TRR was highest in V1, and progressively smaller in V2 and V3 (Figure 5A), whereas temporal variability was similar among the three visual areas (Figure 5D). On the other hand, both temporal variability and response amplitude were modulated by task difficulty and behavioral performance within each visual area (Figure 5A, D). The larger TRRs in V1 (vs. V2 and V3) may arise from some non-neural factor, possibly the unique vasculature of V1, where there are two input arteries, the calcarine artery (branch of the posterior cerebral artery) and posterior temporal artery (branch of the middle cerebral artery). This would support the view that the TRR is purely vascular in origin (Das et al., 2021) and may be modulated (via vascular mechanisms) by neuromodulatory activity arising in the brainstem. It’s also possible that the TRR is actually not smaller across the visual hierarchy, but instead simply appears so because of how the fMRI signal pools across local neural and/or vascular processes. Furthermore, it’s known that some neuromodulatory projections are stronger in some areas than others, meaning that the pooled magnitude of the TRR might be higher in one area than another, even if the local process and its function is the same. For example, the density of noradrenergic (Foote and Morrison, 1987; Shimegi et al., 2016) and cholinergic (Disney et al., 2006) projections differs in V1 and V2.

The TRR may be neural in origin. Our pupillometry results suggest that arousal increases at trial onset in our task. It’s well known that the noradrenergic projections are relatively unspecific (Foote and Morrison, 1987; Aston-Jones and Waterhouse, 2016). If the NE-arousal brainstem system is active at task onset, this would be expected to modulate neural activity in early visual cortex, given prior anatomical and physiological studies (Foote and Morrison, 1987). A number of previous studies have proposed that neuromodulators can alter the weighing of new evidence and prior expectations in cortical computations via changes in synaptic gain (Dayan and Yu, 2006; Heeger, 2017; de Gee et al., 2020; Friston, 2010; Moran et al., 2013), with NE tamping down the effect of the prior. Consistent with this idea, pupil-linked arousal has been found to primarily affect decision biases, which can be modeled as a change in the weighing of prior vs. likelihood. For example, de Gee et al., 2017 and de Gee et al., 2020 found that larger pupil responses were linked with reductions in decision bias (measured via the signal detection theoretical criterion) in a visual detection task and increased activity in a number of neuromodulatory loci, measured using fMRI. Another study manipulated prior expectations on a trial-by-trial basis and found that pupil size dynamically tracked resulting decision biases in human behavior, with smaller pupil responses corresponding to smaller biases (Krishnamurthy et al., 2017). This, taken together with other results showing that pupil-linked arousal also varies with serial choice biases, decision confidence, and prediction errors (Urai et al., 2017; Colizoli et al., 2018), suggests that a portion of variability in behavioral choices may be causally controlled by arousal. Our results, considered from this perspective, may indicate that TRRs have a computational role, shaping neural activity and/or readouts in visual cortex and thereby having a causal effect on behavior. The TRRs we measure with fMRI may therefore reflect the widespread effect of neuromodulators like NE on cortical computation via normal neurovascular coupling (McGinley et al., 2015), separate effects of NE or other neuromodulators on the vasculature (Bekar et al., 2012; Purkayastha and Raven, 2011), or a combination of these factors.

Our findings may also be related to recent neurophysiology studies in animal models linking arousal, global brain states, and visual cortical activity. New recording techniques have allowed researchers to measure the activity of many thousands of neurons at once (Musall et al., 2019; Steinmetz et al., 2019; Stringer et al., 2019). Studies using such techniques have observed widespread responses in visual cortex (and beyond) during delay period and behavioral response times, i.e., in the absence of a stimulus. A number of studies have found that arousal, measured by pupil size, explains much of the variability in this widespread visual cortical activity (Steinmetz et al., 2019; Stringer et al., 2019; Salkoff et al., 2020). Furthermore, in mouse V1, hemodynamic responses recorded during visual stimulation can be decomposed into a local and global component, the latter of which is correlated with changes in pupil size, consistent with the hypothesis that the global component reflects arousal (Pisauro et al., 2016). It remains to be seen if these widespread neural responses give rise to the TRR. We found that the temporal precision of the TRR across trials scaled with arousal level. This is consistent with the hypothesis that the TRR is neural in origin, given previous findings demonstrating that increased alertness leads to more precise neural activity in visual cortex (McGinley et al., 2015; Lombardo et al., 2018; Arazi et al., 2019). If the TRR is neural in origin, it may be driven by a relatively small subpopulation of neurons, not evident in LFPs or mean firing rates (Sirotin and Das, 2009; Das et al., 2021), or by widespread changes in subthreshold neuronal activity (McGinley et al., 2015). The TRR may instead, as described above, be evoked by a widespread release of neuromodulators that both evoke a hemodynamic response and shift the weighing of prior vs. likelihood, both of which may occur without a change in mean firing rates or mean LFP amplitude.

A previous study, Cardoso et al., 2019, found that macaque V1 exhibited alternating increasing and decreasing ramps in blood volume on the timescale of minutes between blocks of low and high reward. This is an interesting result that suggests that the TRR is characterized by both differences in the baseline blood volume and the amplitude of hemodynamic responses. Unfortunately, unlike optical imaging, fMRI-BOLD is not sensitive to such slow changes in the baseline, because the signal is affected by many other sources of noise in the same slow frequency range. Future studies using alternative imaging methods like Vascular-Space-Occupancy (VASO) could potentially resolve these slow ramps in blood volume and validate the results of ref (Cardoso et al., 2019) in humans.

Our findings may be related to previous studies identifying a component of the fMRI-BOLD signal related to arousal, but further research is necessary to test if this component and the TRR reflect the same phenomenon (Chang et al., 2016; Özbay et al., 2019; Gonzalez-Castillo et al., 2021; Goodale et al., 2021). One of these studies Goodale et al., 2021 found that the most responsive voxels measured were in the primary sensory cortices, including early visual cortex, and that these voxels alone sufficed for prediction of behavioral performance and an EEG measure of arousal. If future studies find that this fMRI component is equivalent to the TRR, it could lead to a productive unification of these two areas of research.

The amplitude of the TRR in V1 (mean across all voxels and trials, 0.32% change; median, 0.28% change; max, 6.20% change) is similar to that of stimulus-evoked and attentional hemodynamic responses (Cardoso et al., 2012; Roth et al., 2020; Dugué et al., 2020; Das et al., 2021; Lewis et al., 2018). This raises concerns about fMRI task designs in which attention or other psychological variables are operationalized according to behavioral performance and/or task parameters. In such cases, TRRs, which are modulated by critical task variables and behavior, may mix with other fMRI signal components, and thereby confound analysis. Our results demonstrate that common preprocessing approaches (i.e. global signal regression, physiological ‘noise’ correction) do not remove the TRR from fMRI data. Furthermore, the TRR contains information about brain state that would otherwise typically be measured independently with pupil size and/or cardiac activity. Therefore, in some scenarios it may be useful to model the TRR, remove it from fMRI data as a preprocessing step, and then possibly analyze it as an independent measure of arousal. One way to achieve this is to run a 5–10 min version of a task in which the visual stimulus is confined to one visual field, so that the TRR can be easily isolated and measured later (see Donner et al., 2008 for a simple protocol for removing the TRR).

All data and code are available at osf.io/8fp35 (https://doi.org/10.17605/OSF.IO/8FP35).

1. 1. Burlingham CS
   2. Ryoo M
   3. Roth ZN
   4. Mirbagheri S
   5. Heeger DJ
   6. Merriam EP

   (2022) Open Science Framework

   Task-related hemodynamic responses in human early visual cortex are modulated by task difficulty and behavioral performance.

   https://doi.org/10.17605/OSF.IO/8FP35

1. 1. Arazi A
   2. Yeshurun Y
   3. Dinstein I

   (2019) Neural Variability Is Quenched by Attention

   The Journal of Neuroscience 39:5975–5985.

   https://doi.org/10.1523/JNEUROSCI.0355-19.2019

   • PubMed
   • Google Scholar
2. 1. Arcaro MJ
   2. Honey CJ
   3. Mruczek REB
   4. Kastner S
   5. Hasson U

   (2015) Widespread correlation patterns of fMRI signal across visual cortex reflect eccentricity organization

   eLife 4:e952.

   https://doi.org/10.7554/eLife.03952

   • PubMed
   • Google Scholar
3. 1. Aston-Jones G
   2. Cohen JD

   (2005) An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance

   Annual Review of Neuroscience 28:403–450.

   https://doi.org/10.1146/annurev.neuro.28.061604.135709

   • PubMed
   • Google Scholar
4. 1. Aston-Jones G
   2. Waterhouse B

   (2016) Locus coeruleus: From global projection system to adaptive regulation of behavior

   Brain Research 1645:75–78.

   https://doi.org/10.1016/j.brainres.2016.03.001

   • PubMed
   • Google Scholar
5. 1. Bekar LK
   2. He W
   3. Nedergaard M

   (2008) Locus coeruleus alpha-adrenergic-mediated activation of cortical astrocytes in vivo

   Cerebral Cortex (New York, N.Y 18:2789–2795.

   https://doi.org/10.1093/cercor/bhn040

   • PubMed
   • Google Scholar
6. 1. Bekar LK
   2. Wei HS
   3. Nedergaard M

   (2012) The locus coeruleus-norepinephrine network optimizes coupling of cerebral blood volume with oxygen demand

   Journal of Cerebral Blood Flow and Metabolism 32:2135–2145.

   https://doi.org/10.1038/jcbfm.2012.115

   • PubMed
   • Google Scholar
7. 1. Benjamini Y
   2. Krieger AM
   3. Yekutieli D

   (2006) Adaptive linear step-up procedures that control the false discovery rate

   Biometrika 93:491–507.

   https://doi.org/10.1093/biomet/93.3.491

   • Google Scholar
8. 1. Benson NC
   2. Winawer J

   (2018) Bayesian analysis of retinotopic maps

   eLife 7:224.

   https://doi.org/10.7554/eLife.40224

   • Google Scholar
9. 1. Birn RM
   2. Smith MA
   3. Jones TB
   4. Bandettini PA

   (2008) The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration

   NeuroImage 40:644–654.

   https://doi.org/10.1016/j.neuroimage.2007.11.059

   • PubMed
   • Google Scholar
10. 1. Borovik A
    2. Golubinskaya V
    3. Tarasova O
    4. Aalkjaer C
    5. Nilsson H

    (2005) Phase resetting of arterial vasomotion by burst stimulation of perivascular nerves

    Journal of Vascular Research 42:165–173.

    https://doi.org/10.1159/000084405

    • PubMed
    • Google Scholar
11. Book

    1. Broadbent DE

    (1971)

    Decision and Stress

    Oxford, England: Oxford Press.

    • Google Scholar
12. 1. Burlingham CS
    2. Mirbagheri S
    3. Heeger DJ

    (2022) A unified model of the task-evoked pupil response

    Science Advances 8:eabi9979.

    https://doi.org/10.1126/sciadv.abi9979

    • PubMed
    • Google Scholar
13. 1. Carandini M
    2. Heeger DJ

    (2011) Normalization as a canonical neural computation

    Nature Reviews. Neuroscience 13:51–62.

    https://doi.org/10.1038/nrn3136

    • PubMed
    • Google Scholar
14. 1. Cardoso MMB
    2. Sirotin YB
    3. Lima B
    4. Glushenkova E
    5. Das A

    (2012) The neuroimaging signal is a linear sum of neurally distinct stimulus- and task-related components

    Nature Neuroscience 15:1298–1306.

    https://doi.org/10.1038/nn.3170

    • PubMed
    • Google Scholar
15. 1. Cardoso MMB
    2. Lima B
    3. Sirotin YB
    4. Das A

    (2019) Task-related hemodynamic responses are modulated by reward and task engagement

    PLOS Biology 17:e3000080.

    https://doi.org/10.1371/journal.pbio.3000080

    • PubMed
    • Google Scholar
16. 1. Cazettes F
    2. Reato D
    3. Morais JP
    4. Renart A
    5. Mainen ZF

    (2021) Phasic Activation of Dorsal Raphe Serotonergic Neurons Increases Pupil Size

    Current Biology 31:192–197.

    https://doi.org/10.1016/j.cub.2020.09.090

    • PubMed
    • Google Scholar
17. 1. Chang C
    2. Cunningham JP
    3. Glover GH

    (2009) Influence of heart rate on the BOLD signal: the cardiac response function

    NeuroImage 44:857–869.

    https://doi.org/10.1016/j.neuroimage.2008.09.029

    • PubMed
    • Google Scholar
18. 1. Chang C
    2. Leopold DA
    3. Schölvinck ML
    4. Mandelkow H
    5. Picchioni D
    6. Liu X
    7. Ye FQ
    8. Turchi JN
    9. Duyn JH

    (2016) Tracking brain arousal fluctuations with fMRI

    PNAS 113:4518–4523.

    https://doi.org/10.1073/pnas.1520613113

    • PubMed
    • Google Scholar
19. 1. Chen G
    2. Saad ZS
    3. Britton JC
    4. Pine DS
    5. Cox RW

    (2013) Linear mixed-effects modeling approach to FMRI group analysis

    NeuroImage 73:176–190.

    https://doi.org/10.1016/j.neuroimage.2013.01.047

    • PubMed
    • Google Scholar
20. 1. Chin MS
    2. Kales SN

    (2019) Is There an Optimal Autonomic State for Enhanced Flow and Executive Task Performance?

    Frontiers in Psychology 10:1716.

    https://doi.org/10.3389/fpsyg.2019.01716

    • PubMed
    • Google Scholar
21. 1. Colizoli O
    2. de Gee JW
    3. Urai AE
    4. Donner TH

    (2018) Task-evoked pupil responses reflect internal belief states

    Scientific Reports 8:13702.

    https://doi.org/10.1038/s41598-018-31985-3

    • PubMed
    • Google Scholar
22. 1. Cox RW

    (1996) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages

    Computers and Biomedical Research, an International Journal 29:162–173.

    https://doi.org/10.1006/cbmr.1996.0014

    • PubMed
    • Google Scholar
23. 1. Das A
    2. Murphy K
    3. Drew PJ

    (2021) Rude mechanicals in brain haemodynamics: non-neural actors that influence blood flow

    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 376:20190635.

    https://doi.org/10.1098/rstb.2019.0635

    • PubMed
    • Google Scholar
24. 1. Dayan P
    2. Yu AJ

    (2006) Phasic norepinephrine: A neural interrupt signal for unexpected events

    Network (Bristol, England) 17:335–350.

    https://doi.org/10.1080/09548980601004024

    • PubMed
    • Google Scholar
25. 1. de Gee JW
    2. Knapen T
    3. Donner TH

    (2014) Decision-related pupil dilation reflects upcoming choice and individual bias

    PNAS 111:E618–E625.

    https://doi.org/10.1073/pnas.1317557111

    • PubMed
    • Google Scholar
26. 1. de Gee JW
    2. Colizoli O
    3. Kloosterman NA
    4. Knapen T
    5. Nieuwenhuis S
    6. Donner TH

    (2017) Dynamic modulation of decision biases by brainstem arousal systems

    eLife 6:e23232.

    https://doi.org/10.7554/eLife.23232

    • PubMed
    • Google Scholar
27. 1. de Gee JW
    2. Tsetsos K
    3. Schwabe L
    4. Urai AE
    5. McCormick D
    6. McGinley MJ
    7. Donner TH

    (2020) Pupil-linked phasic arousal predicts a reduction of choice bias across species and decision domains

    eLife 9:e54014.

    https://doi.org/10.7554/eLife.54014

    • PubMed
    • Google Scholar
28. 1. Denison RN
    2. Parker JA
    3. Carrasco M

    (2020) Modeling pupil responses to rapid sequential events

    Behavior Research Methods 52:1991–2007.

    https://doi.org/10.3758/s13428-020-01368-6

    • PubMed
    • Google Scholar
29. 1. Disney AA
    2. Domakonda KV
    3. Aoki C

    (2006) Differential expression of muscarinic acetylcholine receptors across excitatory and inhibitory cells in visual cortical areas V1 and V2 of the macaque monkey

    The Journal of Comparative Neurology 499:49–63.

    https://doi.org/10.1002/cne.21096

    • PubMed
    • Google Scholar
30. 1. Donner TH
    2. Sagi D
    3. Bonneh YS
    4. Heeger DJ

    (2008) Opposite neural signatures of motion-induced blindness in human dorsal and ventral visual cortex

    The Journal of Neuroscience 28:10298–10310.

    https://doi.org/10.1523/JNEUROSCI.2371-08.2008

    • PubMed
    • Google Scholar
31. 1. Duffy S
    2. MacVicar BA

    (1995)

    Adrenergic calcium signaling in astrocyte networks within the hippocampal slice

    The Journal of Neuroscience 15:5535–5550.

    • PubMed
    • Google Scholar
32. 1. Dugué L
    2. Merriam EP
    3. Heeger DJ
    4. Carrasco M

    (2020) Differential impact of endogenous and exogenous attention on activity in human visual cortex

    Scientific Reports 10:21274.

    https://doi.org/10.1038/s41598-020-78172-x

    • PubMed
    • Google Scholar
33. 1. Faber JE

    (1988) In situ analysis of alpha-adrenoceptors on arteriolar and venular smooth muscle in rat skeletal muscle microcirculation

    Circulation Research 62:37–50.

    https://doi.org/10.1161/01.res.62.1.37

    • PubMed
    • Google Scholar
34. 1. Fischl B

    (2012) FreeSurfer

    NeuroImage 62:774–781.

    https://doi.org/10.1016/j.neuroimage.2012.01.021

    • PubMed
    • Google Scholar
35. 1. Foote SL
    2. Morrison JH

    (1987) Extrathalamic modulation of cortical function

    Annual Review of Neuroscience 10:67–95.

    https://doi.org/10.1146/annurev.ne.10.030187.000435

    • PubMed
    • Google Scholar
36. 1. Friston K

    (2010) The free-energy principle: a unified brain theory?

    Nature Reviews. Neuroscience 11:127–138.

    https://doi.org/10.1038/nrn2787

    • PubMed
    • Google Scholar
37. 1. Gandhi SP
    2. Heeger DJ
    3. Boynton GM

    (1999) Spatial attention affects brain activity in human primary visual cortex

    PNAS 96:3314–3319.

    https://doi.org/10.1073/pnas.96.6.3314

    • PubMed
    • Google Scholar
38. Software

    1. Gardner JL

    (2018) MGL: visual psychophysics stimuli and experimental design package

    Zenodo.

    https://doi.org/10.5281/zenodo.1299497
39. 1. Goldberger JJ
    2. Challapalli S
    3. Tung R
    4. Parker MA
    5. Kadish AH

    (2001) Relationship of heart rate variability to parasympathetic effect

    Circulation 103:1977–1983.

    https://doi.org/10.1161/01.cir.103.15.1977

    • PubMed
    • Google Scholar
40. 1. Gonzalez-Castillo J
    2. Panwar P
    3. Buchanan LC
    4. Caballero-Gaudes C
    5. Handwerker DA
    6. Jangraw DC
    7. Zachariou V
    8. Inati S
    9. Roopchansingh V
    10. Derbyshire JA
    11. Bandettini PA

    (2016) Evaluation of multi-echo ICA denoising for task based fMRI studies: Block designs, rapid event-related designs, and cardiac-gated fMRI

    NeuroImage 141:452–468.

    https://doi.org/10.1016/j.neuroimage.2016.07.049

    • PubMed
    • Google Scholar
41. 1. Gonzalez-Castillo J
    2. Fernandez I
    3. Handwerker DA
    4. Bandettini PA

    (2021) The Ubiquitous Vigilance Signal in FMRI Time Series Data

    Neuroscience 1:677.

    https://doi.org/10.1101/2021.07.08.451677

    • Google Scholar
42. 1. Goodale SE
    2. Ahmed N
    3. Zhao C
    4. de Zwart JA
    5. Özbay PS
    6. Picchioni D
    7. Duyn J
    8. Englot DJ
    9. Morgan VL
    10. Chang C

    (2021) fMRI-based detection of alertness predicts behavioral response variability

    eLife 10:e62376.

    https://doi.org/10.7554/eLife.62376

    • PubMed
    • Google Scholar
43. 1. Groppe DM
    2. Urbach TP
    3. Kutas M

    (2011) Mass univariate analysis of event-related brain potentials/fields I: a critical tutorial review

    Psychophysiology 48:1711–1725.

    https://doi.org/10.1111/j.1469-8986.2011.01273.x

    • PubMed
    • Google Scholar
44. 1. Hall CN
    2. Reynell C
    3. Gesslein B
    4. Hamilton NB
    5. Mishra A
    6. Sutherland BA
    7. O’Farrell FM
    8. Buchan AM
    9. Lauritzen M
    10. Attwell D

    (2014) Capillary pericytes regulate cerebral blood flow in health and disease

    Nature 508:55–60.

    https://doi.org/10.1038/nature13165

    • PubMed
    • Google Scholar
45. 1. Heeger DJ
    2. Ress D

    (2002) What does fMRI tell us about neuronal activity?

    Nature Reviews Neuroscience 3:142–151.

    https://doi.org/10.1038/nrn730

    • Google Scholar
46. 1. Heeger DJ

    (2017) Theory of cortical function

    PNAS 114:1773–1782.

    https://doi.org/10.1073/pnas.1619788114

    • PubMed
    • Google Scholar
47. 1. Herman MC
    2. Cardoso MMB
    3. Lima B
    4. Sirotin YB
    5. Das A

    (2017) Simultaneously estimating the task-related and stimulus-evoked components of hemodynamic imaging measurements

    Neurophotonics 4:031223.

    https://doi.org/10.1117/1.NPh.4.3.031223

    • PubMed
    • Google Scholar
48. 1. Herrmann K
    2. Montaser-Kouhsari L
    3. Carrasco M
    4. Heeger DJ

    (2010) When size matters: attention affects performance by contrast or response gain

    Nature Neuroscience 13:1554–1559.

    https://doi.org/10.1038/nn.2669

    • PubMed
    • Google Scholar
49. 1. Jack AI
    2. Shulman GL
    3. Snyder AZ
    4. McAvoy M
    5. Corbetta M

    (2006) Separate modulations of human V1 associated with spatial attention and task structure

    Neuron 51:135–147.

    https://doi.org/10.1016/j.neuron.2006.06.003

    • PubMed
    • Google Scholar
50. 1. Joshi S
    2. Li Y
    3. Kalwani RM
    4. Gold JI

    (2016) Relationships between Pupil Diameter and Neuronal Activity in the Locus Coeruleus, Colliculi, and Cingulate Cortex

    Neuron 89:221–234.

    https://doi.org/10.1016/j.neuron.2015.11.028

    • PubMed
    • Google Scholar
51. 1. Kahneman D
    2. Beatty J

    (1966) Pupil diameter and load on memory

    Science (New York, N.Y.) 154:1583–1585.

    https://doi.org/10.1126/science.154.3756.1583

    • PubMed
    • Google Scholar
52. 1. Kahneman D
    2. Beatty J

    (1967) Pupillary responses in a pitch-discrimination task

    Perception & Psychophysics 2:101–105.

    https://doi.org/10.3758/BF03210302

    • Google Scholar
53. 1. Krishnamurthy K
    2. Nassar MR
    3. Sarode S
    4. Gold JI

    (2017) Arousal-related adjustments of perceptual biases optimize perception in dynamic environments

    Nature Human Behaviour 1:0107.

    https://doi.org/10.1038/s41562-017-0107

    • PubMed
    • Google Scholar
54. 1. Kundu P
    2. Inati SJ
    3. Evans JW
    4. Luh WM
    5. Bandettini PA

    (2012) Differentiating BOLD and non-BOLD signals in fMRI time series using multi-echo EPI

    NeuroImage 60:1759–1770.

    https://doi.org/10.1016/j.neuroimage.2011.12.028

    • PubMed
    • Google Scholar
55. 1. Kundu P
    2. Brenowitz ND
    3. Voon V
    4. Worbe Y
    5. Vértes PE
    6. Inati SJ
    7. Saad ZS
    8. Bandettini PA
    9. Bullmore ET

    (2013) Integrated strategy for improving functional connectivity mapping using multiecho fMRI

    PNAS 110:16187–16192.

    https://doi.org/10.1073/pnas.1301725110

    • PubMed
    • Google Scholar
56. 1. Lewis LD
    2. Setsompop K
    3. Rosen BR
    4. Polimeni JR

    (2018) Stimulus-dependent hemodynamic response timing across the human subcortical-cortical visual pathway identified through high spatiotemporal resolution 7T fMRI

    NeuroImage 181:279–291.

    https://doi.org/10.1016/j.neuroimage.2018.06.056

    • PubMed
    • Google Scholar
57. 1. Lombardo JA
    2. Macellaio MV
    3. Liu B
    4. Palmer SE
    5. Osborne LC

    (2018) State dependence of stimulus-induced variability tuning in macaque MT

    PLOS Computational Biology 14:e1006527.

    https://doi.org/10.1371/journal.pcbi.1006527

    • PubMed
    • Google Scholar
58. 1. Mateo C
    2. Knutsen PM
    3. Tsai PS
    4. Shih AY
    5. Kleinfeld D

    (2017) Entrainment of Arteriole Vasomotor Fluctuations by Neural Activity Is a Basis of Blood-Oxygenation-Level-Dependent “Resting-State” Connectivity

    Neuron 96:936–948.

    https://doi.org/10.1016/j.neuron.2017.10.012

    • PubMed
    • Google Scholar
59. 1. McGinley MJ
    2. David SV
    3. McCormick DA

    (2015) Cortical Membrane Potential Signature of Optimal States for Sensory Signal Detection

    Neuron 87:179–192.

    https://doi.org/10.1016/j.neuron.2015.05.038

    • PubMed
    • Google Scholar
60. 1. Moran RJ
    2. Campo P
    3. Symmonds M
    4. Stephan KE
    5. Dolan RJ
    6. Friston KJ

    (2013) Free energy, precision and learning: the role of cholinergic neuromodulation

    The Journal of Neuroscience 33:8227–8236.

    https://doi.org/10.1523/JNEUROSCI.4255-12.2013

    • PubMed
    • Google Scholar
61. 1. Musall S
    2. Kaufman MT
    3. Juavinett AL
    4. Gluf S
    5. Churchland AK

    (2019) Single-trial neural dynamics are dominated by richly varied movements

    Nature Neuroscience 22:1677–1686.

    https://doi.org/10.1038/s41593-019-0502-4

    • PubMed
    • Google Scholar
62. 1. Nestares O
    2. Heeger DJ

    (2000) Robust multiresolution alignment of MRI brain volumes

    Magnetic Resonance in Medicine 43:705–715.

    https://doi.org/10.1002/(sici)1522-2594(200005)43:5<705::aid-mrm13>3.0.co;2-r

    • PubMed
    • Google Scholar
63. 1. Nilsson H
    2. Aalkjaer C

    (2003) Vasomotion: mechanisms and physiological importance

    Molecular Interventions 3:79–89.

    https://doi.org/10.1124/mi.3.2.79

    • PubMed
    • Google Scholar
64. 1. Özbay PS
    2. Chang C
    3. Picchioni D
    4. Mandelkow H
    5. Chappel-Farley MG
    6. van Gelderen P
    7. de Zwart JA
    8. Duyn J

    (2019) Sympathetic activity contributes to the fMRI signal

    Communications Biology 2:421.

    https://doi.org/10.1038/s42003-019-0659-0

    • PubMed
    • Google Scholar
65. 1. Pan Q
    2. Zhou G
    3. Wang R
    4. Cai G
    5. Yan J
    6. Fang L
    7. Ning G

    (2016) Do the deceleration/acceleration capacities of heart rate reflect cardiac sympathetic or vagal activity? A model study

    Medical & Biological Engineering & Computing 54:1921–1933.

    https://doi.org/10.1007/s11517-016-1486-9

    • PubMed
    • Google Scholar
66. 1. Peppiatt CM
    2. Howarth C
    3. Mobbs P
    4. Attwell D

    (2006) Bidirectional control of CNS capillary diameter by pericytes

    Nature 443:700–704.

    https://doi.org/10.1038/nature05193

    • PubMed
    • Google Scholar
67. 1. Phipson B
    2. Smyth GK

    (2010) Permutation P-values should never be zero: calculating exact P-values when permutations are randomly drawn

    Statistical Applications in Genetics and Molecular Biology 9:Article39.

    https://doi.org/10.2202/1544-6115.1585

    • PubMed
    • Google Scholar
68. 1. Pisauro MA
    2. Benucci A
    3. Carandini M

    (2016) Local and global contributions to hemodynamic activity in mouse cortex

    Journal of Neurophysiology 115:2931–2936.

    https://doi.org/10.1152/jn.00125.2016

    • PubMed
    • Google Scholar
69. 1. Power JD
    2. Plitt M
    3. Gotts SJ
    4. Kundu P
    5. Voon V
    6. Bandettini PA
    7. Martin A

    (2018) Ridding fMRI data of motion-related influences: Removal of signals with distinct spatial and physical bases in multiecho data

    PNAS 115:E2105–E2114.

    https://doi.org/10.1073/pnas.1720985115

    • PubMed
    • Google Scholar
70. 1. Purkayastha S
    2. Raven PB

    (2011) The functional role of the alpha-1 adrenergic receptors in cerebral blood flow regulation

    Indian Journal of Pharmacology 43:502–506.

    https://doi.org/10.4103/0253-7613.84950

    • PubMed
    • Google Scholar
71. 1. Raichle ME
    2. Hartman BK
    3. Eichling JO
    4. Sharpe LG

    (1975) Central noradrenergic regulation of cerebral blood flow and vascular permeability

    PNAS 72:3726–3730.

    https://doi.org/10.1073/pnas.72.9.3726

    • PubMed
    • Google Scholar
72. 1. Ress D
    2. Backus BT
    3. Heeger DJ

    (2000) Activity in primary visual cortex predicts performance in a visual detection task

    Nature Neuroscience 3:940–945.

    https://doi.org/10.1038/78856

    • PubMed
    • Google Scholar
73. 1. Roth ZN
    2. Ryoo M
    3. Merriam EP

    (2020) Task-related activity in human visual cortex

    PLOS Biology 18:11.

    https://doi.org/10.1371/journal.pbio.3000921

    • PubMed
    • Google Scholar
74. 1. Salkoff DB
    2. Zagha E
    3. McCarthy E
    4. McCormick DA

    (2020) Movement and Performance Explain Widespread Cortical Activity in a Visual Detection Task

    Cerebral Cortex (New York, N.Y 30:421–437.

    https://doi.org/10.1093/cercor/bhz206

    • PubMed
    • Google Scholar
75. 1. Shaffer F
    2. Ginsberg JP

    (2017) An Overview of Heart Rate Variability Metrics and Norms

    Frontiers in Public Health 5:258.

    https://doi.org/10.3389/fpubh.2017.00258

    • PubMed
    • Google Scholar
76. 1. Shan ZY
    2. Wright MJ
    3. Thompson PM
    4. McMahon KL
    5. Blokland GGAM
    6. de Zubicaray GI
    7. Martin NG
    8. Vinkhuyzen AAE
    9. Reutens DC

    (2014) Modeling of the hemodynamic responses in block design fMRI studies

    Journal of Cerebral Blood Flow and Metabolism 34:316–324.

    https://doi.org/10.1038/jcbfm.2013.200

    • PubMed
    • Google Scholar
77. 1. Shimegi S
    2. Kimura A
    3. Sato A
    4. Aoyama C
    5. Mizuyama R
    6. Tsunoda K
    7. Ueda F
    8. Araki S
    9. Goya R
    10. Sato H

    (2016) Cholinergic and serotonergic modulation of visual information processing in monkey V1

    Journal of Physiology, Paris 110:44–51.

    https://doi.org/10.1016/j.jphysparis.2016.09.001

    • PubMed
    • Google Scholar
78. 1. Sirotin YB
    2. Das A

    (2009) Anticipatory haemodynamic signals in sensory cortex not predicted by local neuronal activity

    Nature 457:475–479.

    https://doi.org/10.1038/nature07664

    • PubMed
    • Google Scholar
79. 1. Sirotin YB
    2. Cardoso M
    3. Lima B
    4. Das A

    (2012) Spatial homogeneity and task-synchrony of the trial-related hemodynamic signal

    NeuroImage 59:2783–2797.

    https://doi.org/10.1016/j.neuroimage.2011.10.019

    • PubMed
    • Google Scholar
80. 1. Sorriento D
    2. Trimarco B
    3. Iaccarino G

    (2011)

    Adrenergic mechanism in the control of endothelial function

    Translational Medicine @ UniSa 1:213–228.

    • PubMed
    • Google Scholar
81. 1. Stark DE
    2. Margulies DS
    3. Shehzad ZE
    4. Reiss P
    5. Kelly AMC
    6. Uddin LQ
    7. Gee DG
    8. Roy AK
    9. Banich MT
    10. Castellanos FX
    11. Milham MP

    (2008) Regional variation in interhemispheric coordination of intrinsic hemodynamic fluctuations

    The Journal of Neuroscience 28:13754–13764.

    https://doi.org/10.1523/JNEUROSCI.4544-08.2008

    • PubMed
    • Google Scholar
82. 1. Steinmetz NA
    2. Zatka-Haas P
    3. Carandini M
    4. Harris KD

    (2019) Distributed coding of choice, action and engagement across the mouse brain

    Nature 576:266–273.

    https://doi.org/10.1038/s41586-019-1787-x

    • PubMed
    • Google Scholar
83. 1. Stevens JR
    2. Al Masud A
    3. Suyundikov A
    4. Zaykin D

    (2017) A comparison of multiple testing adjustment methods with block-correlation positively-dependent tests

    PLOS ONE 12:e0176124.

    https://doi.org/10.1371/journal.pone.0176124

    • Google Scholar
84. 1. Stringer C
    2. Pachitariu M
    3. Steinmetz N
    4. Reddy CB
    5. Carandini M
    6. Harris KD

    (2019) Spontaneous behaviors drive multidimensional, brainwide activity

    Science (New York, N.Y.) 364:255.

    https://doi.org/10.1126/science.aav7893

    • PubMed
    • Google Scholar
85. 1. Takata N
    2. Yoshida K
    3. Komaki Y
    4. Xu M
    5. Sakai Y
    6. Hikishima K
    7. Mimura M
    8. Okano H
    9. Tanaka KF

    (2015) Optogenetic activation of CA1 pyramidal neurons at the dorsal and ventral hippocampus evokes distinct brain-wide responses revealed by mouse fMRI

    PLOS ONE 10:e0121417.

    https://doi.org/10.1371/journal.pone.0121417

    • PubMed
    • Google Scholar
86. 1. Tootell RB
    2. Hadjikhani N
    3. Hall EK
    4. Marrett S
    5. Vanduffel W
    6. Vaughan JT
    7. Dale AM

    (1998) The retinotopy of visual spatial attention

    Neuron 21:1409–1422.

    https://doi.org/10.1016/s0896-6273(00)80659-5

    • PubMed
    • Google Scholar
87. 1. Urai AE
    2. Braun A
    3. Donner TH

    (2017) Pupil-linked arousal is driven by decision uncertainty and alters serial choice bias

    Nature Communications 8:14637.

    https://doi.org/10.1038/ncomms14637

    • PubMed
    • Google Scholar
88. 1. van den Brink RL
    2. Murphy PR
    3. Nieuwenhuis S

    (2016) Pupil Diameter Tracks Lapses of Attention

    PLOS ONE 11:e0165274.

    https://doi.org/10.1371/journal.pone.0165274

    • PubMed
    • Google Scholar
89. 1. Wang M
    2. He Y
    3. Sejnowski TJ
    4. Yu X

    (2018) Brain-state dependent astrocytic Ca²⁺ signals are coupled to both positive and negative BOLD-fMRI signals

    PNAS 115:E1647–E1656.

    https://doi.org/10.1073/pnas.1711692115

    • PubMed
    • Google Scholar
90. 1. Widdicombe JG
    2. Sterling GM

    (1970) The autonomic nervous system and breathing

    Archives of Internal Medicine 126:311–329.

    https://doi.org/10.1001/archinte.1970.00310080117020

    • PubMed
    • Google Scholar
91. 1. Yarkoni T
    2. Barch DM
    3. Gray JR
    4. Conturo TE
    5. Braver TS

    (2009) BOLD correlates of trial-by-trial reaction time variability in gray and white matter: A multi-study fMRI analysis

    PLOS ONE 4:e4257.

    https://doi.org/10.1371/journal.pone.0004257

    • PubMed
    • Google Scholar

Author details

1. Charlie S Burlingham

   Department of Psychology, New York University, New York, United States

   Contribution

   Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review and editing

   For correspondence

   charlie.burlingham@nyu.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9286-2091

2. Minyoung Ryoo

   Laboratory of Brain and Cognition, National Institute of Mental Health, National Institutes of Health, Bethesda, United States

   Contribution

   Investigation, Writing – review and editing

   Competing interests

   No competing interests declared

3. Zvi N Roth

   Laboratory of Brain and Cognition, National Institute of Mental Health, National Institutes of Health, Bethesda, United States

   Contribution

   Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2173-1625

4. Saghar Mirbagheri

   Graduate Program in Neuroscience, University of Washington, Seattle, United States

   Contribution

   Formal analysis, Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0994-3932

5. David J Heeger

   Department of Psychology and Center for Neural Science, New York University, New York, United States

   Contribution

   Conceptualization, Funding acquisition, Supervision, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-3282-9898

6. Elisha P Merriam

   Laboratory of Brain and Cognition, National Institute of Mental Health, National Institutes of Health, Bethesda, United States

   Contribution

   Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Software, Supervision, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2787-566X

• 1,140

  views

• 180

  downloads

• 11

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.