Neurophysiological markers of alert responding during goal-directed behavior: A high-density electrical mapping study
Introduction
A driver approaches a crossroads, notes the red light and pulls to a stop. Harried and worn out from a fairly dismal day at the office, his eyes are fixed on the lights; his only thoughts are for home, slippers, a Manhattan and an evening in front of the TV. The lights, after what feels like an age, change back to green and he hits his accelerator, relieved to be on his way again. Unfortunately, he has somehow neglected to notice that the car in front of him has yet to pull away. This, he ultimately understands, only as his car's forward progress is all too abruptly arrested by the rear fender of the car in front. Such momentary lapses in attentional re-allocation are unfortunately all too commonplace in life. Avoiding them relies particularly on our ability to sustain attention and to flexibly redeploy our attention to additional environmental factors that might be relevant. This latter function often requires reactivation of a subgoal, outside the immediate spotlight of attention. That is, the primary goal above is to go when the light is green, but an important subgoal is to ensure that there is nothing in the way. Although the example above is of a relatively benign circumstance where a lapse in attention results in an inappropriate response to current environmental circumstances, such lapses can also have catastrophic consequences. For example, a considerable proportion of major road traffic accidents can be attributed to momentary lapses in goal maintenance—that is, distraction or lapses in sustained attentional mechanisms (see, e.g., National survey of distracted and drowsy driving attitudes and behaviors: 20021). As such, it is of great interest to understand the neural mechanisms that support sustained attention and subgoal activation.
Sustained attention requires an intrinsic maintenance of the alert state in the absence of exogenous inputs (Posner and DiGirolamo, 2000, Posner and Peterson, 1990, Sturm et al., 1999). Lapses of sustained attention occur when there is a transient reduction in the alert state that can give rise to momentary loss of endogenous control over behavior. Recent positron emission tomography (PET) studies (Sturm et al., 1999, Sturm et al., 2004) suggest that an extended right hemisphere network is involved in sustained attention including the right anterior cingulate, the right dorsolateral prefrontal cortex, the right inferior parietal lobule with projections to the thalamus and noradrenergic brainstem targets. Sturm and colleagues propose that right hemisphere brain structures exercise top-down control, via the thalamus, on noradrenergic structures in the brainstem.
A task that has proven very effective in assessing this type of attentive responding is the sustained attention to response task (SART) (Manly et al., 1999, Manly et al., 2002, Manly et al., 2003, Robertson et al., 1997). In one version of this task, a predictable series of single digits are presented (1–9) and subjects are required to make a response to each number (go trials) with the exception of the number 3 (no-go trial). A PET study showed that this task increased activation in both the right dorsolateral prefrontal cortex and the right superior/posterior parietal cortex compared to a more challenging version of the SART in which the numbers were presented randomly (Manly et al., 2003). These findings suggest that right fronto-parietal regions are responsible for maintaining a goal-directed focus in unarousing contexts where exogenous stimuli are not present to increase alertness through novelty, demand or perceived difficulty (Robertson and Garavan, 2004).
The SART has also proven to be a sensitive clinical measure, discriminating between traumatically brain injured (TBI) patients and healthy control subjects (Dockree et al., 2004, Manly et al., 2003, McAvinue et al., 2005, Robertson et al., 1997) and between ADHD children and controls (Shallice et al., 2002). Clinical groups exhibit increased errors of commission (false presses on the 3) during this task and fail to show anticipatory slowing on the trials before the upcoming no-go trial, a general finding in subjects who are successful at this task (Dockree et al., 2004), suggesting a loss of endogenous control at strategically important points during the task. Although, the functional anatomical correlates of the SART have been investigated (Fassbender et al., 2004, Manly et al., 2003, O'Connor et al., 2004), only one study (Dockree et al., 2004) has examined the electrophysiological dynamics during the task. In the latter study, alpha (∼10 Hz) desynchronization was observed in healthy controls before the no-go trial. By contrast, TBI patients failed to show this modulation. This state of desynchronization has been associated with increased attentive processing (Klimesch et al., 1998, Mulholland, 1965, Pfurtscheller and Lopes da Silva, 1999) in the transition from a relaxed to an alert state, and with anticipatory preparation of visual cortices during selective attention tasks (Foxe et al., 1998, Fu et al., 2001, Worden et al., 2000). No studies to date have characterized the broad-band ERP componentry of the SART in neurologically normal adults.
In the present study, we utilize the excellent temporal resolution of high-density electrical mapping to examine the spatiotemporal dynamics of electro-cortical activity during the fixed sequence SART (hereafter referred to as the SARTfixed). Our first aim was to examine event-related potentials (ERPs) and their topographical distributions during periods of accurate sustained attention performance described as ‘successful runs’ of trials 1 through 9 (i.e., when subjects successfully withhold responses to trial 3). In previous investigations of the SARTfixed, adequate characterization of errors of commission has not been possible due to the rarity of their occurrence. In this study we addressed this limitation by testing subjects over the course of a full day (with regular breaks) so that adequate numbers of errors were committed. A criticism of this approach pertains to the ecological validity of the task. Can long-term engagement with a laboratory task over ∼108 min relate to a more naturalistic situation? We argue that the task has features in common with everyday scenarios that require this kind of sustained attentional effort over long periods. For example, taking a long distance trip in the car interspersed with regular breaks may require similar periods of sustained attention to critical events during largely routine behavior. The SART also correlates with everyday reported cognitive failures in patients with traumatic brain injury (Robertson et al., 1997) suggesting that the propensity for attentional lapses on the task is related to greater everyday slips of attention.
Accordingly, as a second aim, we conducted an exploratory analysis of correct withholds and commission errors as well as the trials that preceded and followed these responses. The current study, in neurologically healthy subjects, will provide an important baseline for future understanding of clinical populations and their documented sustained attention deficits on this task. We acknowledge that we would not be able to test clinical groups for long periods of time due to excessive fatigue. However, in these groups the number of errors mounts up much more rapidly, circumventing the need to test for long periods.
We outline a number of predictions regarding the ERP componentry during the critical anticipatory period before the upcoming no-go target that will serve as important markers for alert responding during the task. On trial 2, we predict that early visual attentional processes will be mobilized because of the significance of this trial as an upcoming cue for the critical target trial. The most commonly reported attentional modulation is that of the P1 and N1 components that show increased amplitudes when spatial attention is directed to the stimulus location (i.e., stimuli are validly cued) compared to when it is directed elsewhere (Mangun and Hillyard, 1991, Mangun et al., 1987). In contrast, selectively attending to relevant visual features has been shown to elicit the so-called selection negativity (SN) (Anllo-Vento and Hillyard, 1996, Harter et al., 1984, Kenemans et al., 1993, Smid et al., 1999, Molholm et al., 2004). Although the SARTfixed is not a selective visual feature attention paradigm per se, it is reasonable to propose that similar processes of visual selection might be engaged as the relevance of trials in the sequence increase before the critical target trial. Indeed, previous work has demonstrated that increased ventral-stream visual object-recognition processes underlying SN can be elicited by relevant visual stimuli when relevance is defined on the basis of a non-spatial feature(s) (Anllo-Vento and Hillyard, 1996, Harter et al., 1984, Kenemans et al., 1993, Smid et al., 1999).
We further predict that the recollection of the task goal (“withhold response to 3”) will be critical on trial 2. Research investigating the dynamics of prospective remembering (West et al., 2001, West and Krompinger, 2005) has shown that the realization of a delayed intention is associated with two ERP modulations: an N300 and a “prospective positivity”. They propose that the N300 is associated with detection of prospective memory cues and is seen as a phasic negativity over occipito-parietal scalp between 300 and 500 ms. Additionally, the later prospective positivity, seen as a broadly distributed positivity (500–1000 ms) over parietal areas, may reflect neural processes supporting the recollection of a task goal. Interestingly, the later phase of the prospective positivity (600–800 ms) is not influenced by the salience of the prospective cue suggesting that the prospective positivity is elicited under different circumstances than the more commonly observed parietal P3b—a component that is observed over parietal scalp in the context of target detection and that is augmented by target distinctiveness (Comerchero and Polich, 1998).
From trials 2 to 3, subjects must switch between a go response to a withhold response. Wylie et al. (2003) have examined task switching during a paradigm in which subjects regularly alternate between two tasks (categorizing numbers and categorizing letters) on every third trial. On the preceding trial to a switch trial, a period of sustained positivity was observed over bilateral parietal regions. This effect was interpreted in the context of a “competition” model in which either the currently relevant task set is suppressed or the subsequent trial task set is activated, or both. It is probable that competition needs to be resolved in the context of different task sets during the SARTfixed and there will be sustained positivity that is indicative of this competition in the transition between responses on trials 2–3.
Two possibilities can be proposed regarding the sensory processing that the target (trial 3) stimulus might receive. One possibility is that since this stimulus is specifically not to be responded to, sensory processing of trial 3 might actually be suppressed, with visual attentional resources being withdrawn from the stimulus. The second, and perhaps more plausible prediction, is that trial 3 will be processed in a state of high visual attentiveness so that subsequent subgoal engagement can occur. These two possibilities lead to distinct predictions regarding the sensory-evoked activity that will be generated to trial 3. If processing of trial 3 is selectively suppressed, then the early P1–N1 complex of the VEP should be reduced in amplitude. Alternately, if trial 3 receives extra visual attentional processing, then the pattern of activity during sensory processing should parallel that seen during trial 2; that is, we should see attentional enhancement and an SN. Further, we also expect that trial 3 will result in generation of the N2–P3 complex that has frequently been observed during the presentation of no-go stimuli (Bekker et al., 2004, Kok, 1986, Pfefferbaum et al., 1985). The N2 component has been interpreted as reflecting conflict monitoring during infrequent trials whereas the P3 is more associated with response inhibition or evaluation of an erroneous response after response execution (Kok et al., 2004).
In addition to evaluating the predictions above, we performed a second exploratory analysis as a means of fully characterizing the richness of our data set and as a hypothesis generation tool for future research. We asked whether there are different ERP time courses prior to and during a correct withhold versus prior to and during a commission error.
Section snippets
Subjects
Fourteen (six female) right-handed neurologically normal volunteers participated. They were paid $100 for 1 day of participation. Subjects were aged between 18 and 32 years (mean = 23.86, SD ± 4.24). All subjects gave written informed consent, and the Institutional Review Board of the Nathan Kline Institute approved the procedures. All subjects reported normal or corrected-to-normal vision.
SART paradigm and procedure
Digits were presented sequentially from ‘1’ through ‘9’. For each block, 225 digits were presented
Behavioral results
Errors committed during the task were categorized as commission errors defined as a false press on the target trial (‘3’). We adopted conservative inclusion criteria for commission errors and rejected those that were preceded by or followed by a non-response, reasoning that these non-genuine commission errors may reflect an early or late go trial response. Owing to the different number of blocks completed by subjects, percentage error scores were calculated. Mean percentage commission
Discussion
In this experiment we have examined response time patterns and the spatiotemporal dynamics of ERPs that characterize sustained attention performance in neurologically healthy subjects. Behaviorally, a distinct pattern of responses characterized task performance demonstrating a lengthening of RTs at the turnover of the digit sequence from trials 9 to 1 and a subsequent shortening of RTs from trials 1 to 2 prior to the upcoming target on trial 3. Electrophysiologically, we identified a number of
Summary of ERP componentry
Distinct differences in the ERP componentry separated a critical target processing period (trials 9–3) and a non-critical ‘task-driven’ period with relatively lower attentional demands (trials 5–8). We used a collapsed average of the latter as a reference waveform (REFW) to which the active attentional demands during trials 9–3 could be compared. The first differences in ERP amplitude were observed between trials 9 and 1 with a reduction in the amplitude of the P2 component over
Utilizing the number sequence
The data show that the turnover of the digit sequence from trials 9 to 1 results in a lengthening of RTs implying that this transition may be a critical juncture during which stimulus processing is enhanced. In support of this conjecture, our findings demonstrate that P450 amplitude increased over parieto-central regions from trial 9 to trial 1 (see Fig. 5). Our later exploratory analysis revealed that this P450 effect was present to the same magnitude prior to a correct withhold and an error
Selection negativity and no-go negativity
Trials 2 and 3 were exceptional in terms of the early sensory processing they received. We proposed that early visual attentional processes would be mobilized during trial 2 because of its significance as an upcoming cue trial. Further, we suggested that trial 3 might also elicit the same visual attentional processes given its significance as the target trial. However, another possibility was that trial 3 might actually receive less processing due to a suppression of early sensory activity in
Goal recollection
We proposed that recollection of the task goal would be associated with phasic negativity over occipito-parietal scalp and positivity over parietal areas during trial 2, in keeping with previous reports (West and Krompinger, 2005, West et al., 2001). Our findings demonstrate two significant modulations occurred during trial 2: an earlier N2-like effect over occipito-parietal scalp and a later positive (LP1) modulation (550–800 ms) over occipito-parietal and central sites. Both components were
Task switching
We predicted that in the transition from trials 2 to 3 sustained positivity over parietal regions would be indicative of a switch from the go response to the withhold response. Between 850 and 1150 ms, amplitude markedly increased on trial 2 across occipito-parietal and central sites before the presentation of trial 3. In accordance with Wylie et al. (2003), we argue that the increase in the second late positive (LP2) amplitude reflects increased competition between two responses: either the
Fronto-polar positivity
Distinct bilateral anterior fronto-polar positive foci were highly prominent on each trial in the SART sequence. Involvement of fronto-polar cortices in the SART task is to be expected as numerous functional imaging studies have now implicated fronto-polar cortices as critical nodes in the circuit for maintaining and generating subgoals and for integrating such subgoals with the ongoing primary task (e.g., Braver and Bongiolatti, 2002, Badre and Wagner, 2004). In the present task, the main goal
Lapses of attention: failure of goal-directed behavior
In this investigation, the high number of experimental blocks undertaken by subjects ensured that an adequate number of errors were committed over the course of testing to obtain reliable individually averaged waveforms for errors of commission and for the trials preceding these attentional lapses. The most notable difference was a clear divergence of the LP1 component, which exhibited greater amplitude on trial 2 prior to a correct withhold compared to before a commission error—the difference
Conclusion
Here, we characterize the spatiotemporal dynamics of alert responding during an established task that measures one's ability to maintain a goal-directed focus during unarousing conditions. Neurophysiological results reported here inform us about several important processes that reflect alert responding and the failure of goal-directed behavior. These include the mobilization of early visual processing, cue selection, goal recollection, task switching, error evaluation and fronto-polar
Acknowledgments
We would like to thank Manuel Gomez-Ramirez, Beth Higgins, Jeannette Piesco and Marina Shpaner for their excellent technical support. This work was supported in part by grants from the National Institute of Mental Health (MH63434 and MH65350) and the National Institute on Aging (AG22696) to Dr. J.J. Foxe. An International Collaboration Travel Support grant from Enterprise Ireland supported S.P. Kelly's work in New York. A Fulbright Award supported Dr. R.B. Reilly's work in New York. We are also
References (57)
- et al.
Selection, integration, and conflict monitoring: assessing the nature and generality of prefrontal cognitive control mechanisms
Neuron
(2004) - et al.
Electrophysiological correlates of attention, inhibition, sensitivity and bias in a continuous performance task
Clin. Neurophysiol.
(2004) - et al.
The role of frontopolar cortex in subgoal processing during working memory
NeuroImage
(2002) - et al.
Response priming in a go/nogo task: do we have to explain the go/nogo N2 effect in terms of response activation instead of inhibition?
Clin. Neurophysiol.
(2001) - et al.
P3a, perceptual distinctiveness, and stimulus modality
Cogn. Brain Res.
(1998) - et al.
Abstract representations of numbers in the animal and human brain
Trends Neurosci.
(1998) - et al.
EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis
J. Neurosci. Methods
(2004) - et al.
Behavioural and physiological impairments of sustained attention after traumatic brain injury
Brain Res. Cogn. Brain Res.
(2004) - et al.
The roles of timing and task order during task switching
NeuroImage
(2002) Effects of attention and stimulus probability on ERPs in a go/no-go task
Biol. Psychol.
(1993)
A topography of executive functions and their interactions revealed by functional magnetic resonance imaging
Brain Res. Cogn. Brain Res.
Attention-dependent suppression of distracter visual input can be cross-modally cued as indexed by anticipatory parieto-occipital alpha-band oscillations
Brain Res. Cogn. Brain Res.
Event-related potentials to conjunctions of spatial frequency and orientation as a function of stimulus parameters and response requirements
Electroencephalogr. Clin. Neurophysiol.
Induced and evoked band power changes in an oddball task
Electroencephalogr. Clin. Neurophysiol.
Effects of degradation of visual stimulation on components of the event-related potential (ERP) in go/nogo reaction tasks
Biol. Psychol.
Brain responses to number sequences with and without active task requirement
Clin. Neurophysiol.
The absent mind: further investigation of sustained attention to response
Neuropsychologia
Coffee in the cornflakes: time-of-day as a modulator of executive response control
Neuropsychologia
Visuo-spatial neural response interactions in early cortical processing during a simple reaction time task: a high-density electrical mapping study
Neuropsychologia
ERPs to response production and inhibition
Electroencephalogr. Clin. Neurophysiol.
Event-related EEG/MEG synchronization and desynchronization: basic principles
Clin. Neurophysiol.
Impact of instructed relevance on the visual ERP
Int. J. Psychophysiol.
‘Oops!’: performance correlates of everyday attentional failures in traumatic brain injured and normal subjects
Neuropsychologia
EEG alpha power changes reflect response inhibition deficits after traumatic brain injury (TBI) in humans
Neurosci. Lett.
Functional anatomy of intrinsic alertness: evidence for a fronto-parietal-thalamic-brainstem network in the right hemisphere
Neuropsychologia
Network for auditory intrinsic alertness: a PET study
Neuropsychologia
Neural correlates of prospective and retrospective memory
Neuropsychologia
Neural activity associated with the realization of a delayed intention
Cogn. Brain Res.
Cited by (55)
Exploring the sustained and divided attention of novice versus experienced drivers
2022, Transportation Research Interdisciplinary PerspectivesAltered Brain Network Dynamics in Schizophrenia: A Cognitive Electroencephalography Study
2018, Biological Psychiatry: Cognitive Neuroscience and NeuroimagingSustained attention ability in schizophrenia: Investigation of conflict monitoring mechanisms
2017, Clinical NeurophysiologyCitation Excerpt :ISI were 1500 ms, 1700 ms, 2100 ms, 2300 ms, 2500 ms. Five randomly allocated digit sizes were presented to extend the demand for processing the numerical value and to diminish the possibility that subjects would set a search template for some peripheral feature of the target digit (Robertson et al., 1997; Dockree et al., 2005). Digit Arial font sizes were 100, 120, 140, 160 and 180 (Dockree et al., 2005). The five allocated digit sizes subtended vertical angles of 1.39°, 1.66°, 1.92°, 2.18° and 2.45°, respectively, at a viewing distance of 70 cm.