Functional brain imaging of a complex navigation task following one night of total sleep deprivation: a preliminary study


Gary Strangman, Neural Systems Group, 149 13th Street – Psychiatry – Rm 2651, Charlestown, MA 02129, USA. Tel.: +1 617 724 0662; fax: +1 617 726 4078; e-mail:


Several neuroimaging studies have demonstrated compensatory cerebral responses consequent to sleep deprivation (SD), but all have focused on simple tasks with limited behavioral response options. We assessed the cerebral effects associated with SD during the performance of a complex, open-ended, dual-joystick, 3D navigation task (simulated orbital docking) in a cross-over protocol, with counterbalanced orders of normal sleep (NS) and a single night of total SD (∼27 h). Behavioral performance on multiple measures was comparable in the two sleep conditions. Functional magnetic resonance imaging revealed multiple compensatory SD > NS cerebral responses, including the posterior superior temporal sulcus [Brodmann area (BA) 39/22/37], prefrontal cortex (BA 9), lateral temporal cortex (BA 22/21), and right substantia nigra. Right posterior cingulate cortex (BA 31) exhibited NS > SD activity. Our findings extend the compensatory cerebral response hypothesis to complex, open-ended tasks.


A lack of sleep has been demonstrated to produce performance deficits in experimental tasks of alertness, attention, memory, cognition, learning, and motor responses (Harrison and Horne, 2000). While the neurophysiological effects of sleep deprivation (SD) remain incompletely understood, a few recent studies have begun to provide guidance on where and what neurophysiological changes occur as a function of SD (Bell-McGinty et al., 2004; Chee and Choo, 2004; Drummond et al., 1999, 2000, 2001, 2004; Portas et al., 1998; Thomas et al., 2000). Collectively, these studies have demonstrated SD-dependent neurophysiological responses, both globally and locally, for single- and multi-night SD. The particular brain regions affected by SD, however, tend to depend on the task performed (Drummond and Brown, 2001).

Tasks used to date in SD neuroimaging have been relatively simple, requiring only limited behavioral responses. Tasks of concern with respect to SD, in contrast, tend to be complex, challenging, and open-ended (i.e. multiple behavioral response options, including sensorimotor navigation tasks such as driving or flying). Such tasks are commonly performed by persons with sleep deficits; they can give rise to performance failures, and such failures can have serious or fatal consequences. At the same time, complex tasks tend to be less behaviorally impaired by SD than simpler tasks (Gillberg and Åkerstedt, 1998; Harrison and Horne, 2000). The reason for this performance preservation remains unclear, but it has been suggested that performance levels may be more easily maintained when multiple options exist for compensatory behavior. To our knowledge no complex, real-world type task has been employed while investigating the neurophysiological correlates of SD.

The purpose of this study, therefore, was to examine the cerebral activity changes associated with a single night of SD on a complex, open-ended, 3D sensorimotor navigation task (proximal operations associated with docking two spacecraft in orbit). Such docking maneuvers are regularly performed onboard spacecraft, and are exceptionally high-risk endeavors. Moreover, mission demands and altered light/dark cycles lead to particularly disrupted sleep patterns (Wyatt et al., 1999), making SD highly relevant to such a task. Docking also shares multiple attributes with other tasks often performed under SD conditions (e.g. commercial aviation, fighter jet aviation, and driving): all require vigilance, manipulation of multiple actuators, and spatiotemporally precise motor performance. Based on the existing literature, we hypothesized an adaptive cerebral response following SD, whether or not performance deficits were in evidence. (A second goal of this study was to compare the ability of functional magnetic resonance imaging (fMRI) and near infrared imaging to detect such effects. Here we describe the fMRI findings only.)



Six healthy, right-handed volunteers (all non-pilots naïve to the docking task) provided informed consent, and completed the entire protocol (four males, two females; 20 ± 1 (mean ± stdev) years). Additional subjects could not be scanned due to the replacement of MRI hardware. The study was approved by the Institutional Review Board at Massachusetts General Hospital.

Experimental protocol

The experiment lasted for 2 weeks, during which subjects kept a nightly sleep journal and were asked to maintain their normal sleep (NS) schedule and caffeine intake, but refrain from alcohol or drug use. On either day 4 or 5, each subject was pretrained to a fixed criterion performance on the docking task (accurate performance on eight consecutive docking trials, 15–30 min of practice). Brain imaging was performed on days 7 and 14, beginning at either 10:00 or 12:00 hours (the same time of the day when re-scanning a given subject). During scanning, each subject performed five runs of the task. One scanning session followed a night of NS (range 6.5–9 h), and the other followed a night of total SD in the laboratory (continuous awake state of 27 ± 1 h, verified via continuous, direct monitoring by a lab technician), in counterbalanced order.

fMR brain imaging protocol

MRI scanning utilized a Siemens Sonata 1.5 T scanner (Siemens AG, Munich, Germany), collecting one 3D MP-RAGE structural scan (TR = 7.25 ms, TE = 3 ms, flip = 7 degrees, 192 × 256 voxels in 256 mm FOV) and five blood oxygenation level dependent (BOLD) echoplanar functional scans (TR = 2.0 s, TE = 40 ms, 64 × 64 voxels, 3.125 × 3.215 in-plane resolution, twenty 5 mm slices, and 0.75 mm skip) per session. In the scanner, subjects viewed a projected computer display and held two joysticks, one beside each leg. One experimental run is shown in Fig. 1 and consisted of four active periods (48 s each comprised of two 24 s docking trials) interleaved with five fixation periods (24 s each), during which a total of 156 fMRI volumes were acquired. During each docking approach (‘trial’), subjects performed a custom, 3D sensorimotor navigation task relevant to spaceflight, named SpaceDock. The dual-joystick task (x/y control in the right hand, pitch/roll control in the left) required flying a remote control spacecraft to a center-screen docking position. Joystick contacts produced small, fixed increments or decrements in spacecraft velocity. Eight equidistant starting positions were used, and the starting attitude was set to a 25-degree departure from docking-nominal in roll (four trials per run) or pitch (the remaining four trials). Immediately following the moment of contact, performance was evaluated via a 2 s thermometer-style display as a weighted combination of final position and attitude error. Behavioral data was recorded to enable reconstruction of the entire trajectory taken by the subject.

Figure 1.

Schematic of experimental paradigm. Eight dual-joystick, 3D navigation task trials (SpaceDock ‘approaches’) were performed per run, in four blocks of two trials each. Task periods were alternated with a fixation task. Each subject performed five such runs following normal sleep (NS) and five more runs following ∼27 h of total sleep deprivation (SD).

Data analysis

Behavioral data was analyzed for five distinct, flight-relevant performance parameters: (1) distance and (2) angle error at point of contact, (3) composite score (the unitless combination of (1) and (2) used for evaluation), (4) total distance traveled, and (5) number of thruster inputs (i.e. ‘fuel use’). All parameters except angle error were log-normalized. Within-subject tests between the NS and SD conditions utilized repeated-measure anovas, with Run as the random variable. Between-subject analyses consisted of repeated-measures Order × Sleep Condition × Run (2 × 2 × 5) anovas using Subject as the random variable. In both within- and between- subject analyses, each subject contributed five measurements (one per run, each being an average of eight trials) for each sleep condition on each behavioral measure.

Functional magnetic resonance imaging data processing proceeded as follows: (1) within-session motion correction, (2) session co-registration, (3) spatial smoothing (Gaussian kernel, 8 mm full width at half-maximum), (4) piecewise-linear detrending between rest periods (cf. Marchini and Ripley, 2000), (5) multiplicative normalization of each voxel time series to fix the mean fixation fMRI amplitude at 1000, (6) averaging all five runs for each subject in each condition, (7) Talairach resampling, and (8) subject-by-subject statistical mapping using voxel-wise Student's T-tests (no correction for time series autocorrelation). All steps were performed with AFNI v2.51a (Cox, 1996), with the exception of steps 4 and 5 which employed in-house software.

To address the limitation of our small subject pool, we focused solely on voxels activated consistently across subjects. To do so, Talairach-space statistical maps from each subject were first thresholded at a modestly strict, uncorrected P < 0.0001. For task versus fixation analyses, voxels were retained if and only if six of six subjects exhibited significant increases (decreases) in that voxel. When performing the stricter SD versus NS comparison, we relaxed this criterion to five or more of our six subjects. These n-of-six criteria help limit Type I error associated with individual response patterns. The resulting activation maps were then pruned to leave only activation clusters exceeding five voxels (250 mL) in size, thereby reducing Type I error from multiple spatial comparisons.


Subjects slept for 7 ± 1 h (mean ± stdev) per night throughout the study. During postexperiment debriefing, all subjects indicated that the task performance seemed more effortful under SD relative to NS conditions. In addition, no subject was conscious of a change in the performance strategies between conditions. Few consistent performance differences were observed between conditions, either within or between subjects. Between-subject analysis showed one significant interaction between Order and Sleep Condition in the composite (feedback) score, F(1,4) = 17.0, P = 0.015. This reflected slightly improved performance on the second day of scanning, regardless of condition. Only three trends were in evidence (Order by Sleep Condition for distance and angle error, and Run for thruster inputs, P < 0.10 in all three cases). Within-subjects, 11 of the 30 possible effects (six subjects × five variables) were significant, but never in consistent directions (six significant effects were NS > SD, five showed SD > NS; and, whenever one person showed significant NS > SD performance for a given variable, another showed SD > NS). Thus, although performance was variable across subjects – as expected for a complex, open-ended task – overall task performance was roughly equated across sleep conditions.

Neuroimaging comparisons of docking versus fixation revealed similar overall BOLD activation patterns in NS and SD (data not shown). Regions significantly activated in both conditions included left primary sensorimotor and premotor cortex [Brodmann areas (BAs) 6, 4], left postcentral gyrus and precuneus (BA 40/7), and bilateral middle occipital gyrus (BA 19/37). In addition, NS revealed reliable decreases in activity in the posterior cingulate (BA 31), whereas SD instead revealed reliable activity increases relative to fixation in right frontal cortex (BA 9, 46). Collapsing across the two conditions revealed activity in all these regions, plus regions of deactivation in medial prefrontal cortex (bilateral BA 10) and left superior temporal gyrus (BA 21/22).

The direct comparison of docking periods in SD and NS conditions appears in Fig. 2 and Table 1. We observed predominantly greater activation in the SD condition relative to NS (i.e. SD > NS; details in Table 1). There was one notable exception; the right posterior cingulate cortex (BA 31; Z = +24 mm) exhibited NS > SD activity.

Figure 2.

Cerebral correlates of SpaceDock performance following sleep deprivation (SD) versus task performance following normal sleep (NS). Red indicates SD > NS and blue indicates NS > SD; uncorrected P-values. Numbers correspond to the numbered regions in Table 1.

Table 1.  Regions exhibiting activity differences during SpaceDock task performance between sleep-deprived and normal-sleep conditions (item 4 reflects blue in Fig. 2)
RegionHemiBA Vol (mm3)TX(mm)Y(mm)Z(mm)SD versus Fix.*NS versus Fix.* nEffect size (Cohen'sd)
  1. BA, Brodmann area; SD, sleep deprivation; NS, normal sleep; Fix., fixation.

  2. *Columns indicating whether cerebral activity during the task was greater than (>), less than (<) or equal to (=) the cerebral activity observed during fixation.

  3. Regions were reported only when they exhibited significant differences in either five or six out of six subjects.

1 Middle frontal gyrusR932417.341542>=5/60.59
2 Middle temporal gyrusR2159435.156−533>=5/60.20
3 Angular gyrusR3927031.735−5933><5/61.19
4 Posterior cingulateR31567−27.323−6224=>5/60.43
5 Middle temporal gyrusL39/22/37191732.0−38−5918>=6/60.44
6 Middle temporal gyrusR3929728.247−5915>=5/60.44
7 Substantia nigra/pyramidal tractR35135.98−11−13><5/62.02

Fig. 3 displays time courses for each sleep condition (averaged across all subjects) from four selected brain reigons, to investigate the source of activations in Fig. 2. Fig. 3a–c reflect SD > NS cases. The left posterior superior temporal sulcus (pSTS; BA 37) exhibited increases relative to rest in both conditions, with larger increases following SD (‘upregulation’). In contrast, more superior right BA 39 exhibited significant (and possibly growing) decreases in the NS condition, with no apparent modulation in the SD condition (‘downregulation’). In the right substantia nigra, increases were observed after SD, but not after NS (‘recruitment’). In contrast, Fig. 3d reflects the NS > SD case, wherein right BA 31 exhibited positive modulations after NS, but not following SD (‘dropout’).

Figure 3.

Four distinct patterns of average blood oxygenation level dependent (BOLD) modulation following sleep deprivation (SD) versus normal sleep (NS) (cf. Fig. 2). (a) Mean percent signal change for SD (light trace) and NS (heavy trace) for the voxel in left Brodmann area (BA) 37 exhibiting the maximum difference between conditions. Shaded regions indicate periods of task performance. Note the larger task-related BOLD increases in the SD condition. (b) Right BA 39 revealed strong (and possibly growing) signal decreases during task performance following NS, but a tendency toward signal increases following SD. (c) The right substantia nigra exhibited task-related increases following SD but a tendency towards decreases following NS. (d) Right BA 31 exhibited the reverse pattern from (c).


We present the first investigation of the cerebral correlates of SD in a complex, open-ended task. Based on previous SD brain imaging in more standardized laboratory tasks, we hypothesized an adaptive cerebral response following SD. We observed significant activity differences in a meaningful network of brain regions (see below) during SD relative to NS, thereby extending support for the adaptive cerebral response hypothesis to the domain of complex, open-ended tasks.

Cortical observations

The SD-related cortical modulations revealed a coherent network of areas involved in a compensatory cerebral response to SD. The most robust cortical finding, observed in six of six subjects, was a large area of SD > NS activity around the left pSTS (the junction of BAs 39, 22 and 37). This region was also identified by the globally sensitive cerebral metabolic rate for glucose (CMRglu) study of Thomas et al. (2000) as well as a recent fMRI study (Drummond et al., 2004). The pSTS has been associated with non-biological, rigid motion inter-relationships as well as cognitive planning or simulation of complex motion (Cohen et al., 1996; Dukelow et al., 2001). We also observed dorsolateral prefrontal cortex (DLPFC) activity (including BA 9), which has been implicated in the working and episodic memory processing and cognitive control (Buckner et al., 1999), consistent with remembering how to operate the joystick controls, and expectations of the visual display's response to those controls. Activity in the right temporal lobe (BA 21) has been associated with general visuospatial processing, and selective spatial attention in particular (Waberski et al., 2002). Right BA 39 around the angular gyurs has also been associated with spatial attention (Chambers et al., 2004). Together, these cortical regions, each exhibiting SD > NS activity, support visuospatial processing, memory and attention – all conceptually important components of the docking task.

One unique finding was in the posterior cingulate (BA 31). Multiple neuroimaging studies of mild cognitive impairment and dementia have shown that reduced blood flow in this region is an early indicator of cognitive decline (Devous, 2002; Wolf et al., 2003). Here, we observed significant modulation of this region following NS but none (or perhaps activity decreases relative to fixation; Fig. 3d) following SD. Thus, we tentatively propose that the loss of task-related BA 31 modulation following SD might directly reflect declining cognitive function. This remains to be investigated in more detail.

Parietal cortex and SD

Over a series of three studies, Drummond and Brown (2001) previously proposed that SD induces a switch from temporal to parietal processing, at least in verbal learning and divided attention tasks, further suggesting that engagement of parietal regions represented a less efficient and less accurate method of performing these tasks. We did not find evidence to support this hypothesis in our docking task. Moreover, the time course data (Fig. 3) suggests that the changes in cerebral utilization following SD are considerably more complex than a simple ‘switch’ from activating one network to activating another.

However, in a more recent paper specifically investigating task difficulty, Drummond et al. (2004) found inferior parietal cortex significantly associated with successful completion of difficult variants of Baddeley's Grammatical Transformation (GT) task. This finding appeared in the absence of behavioral differences between SD and NS, and was located near the pSTS, the site of our most robust finding. The GT task, however, is quite different from SpaceDock: GT is event-oriented, with verbal cues, keypress responses, and requires substantial working memory, whereas SpaceDock is continuous, with implicit visuospatial cues, complex sensorimotor response requirements, and arguably depends less on working memory. The tasks share two characteristics, however: both have a significant spatial component, and both are challenging. From just these two studies, one cannot attribute activity in the pSTS specifically to spatial- or difficulty-related causes. However, the replicated finding in two highly distinct tasks highlights Drummond et al.'s suggestion that more focus should be placed on the parietal lobes in overcoming the effects of sleep loss.

Other sources of SD versus NS differences

A number of alternative, (non-exclusive) hypotheses can be put forth to explain the observed differences in cerebral activity between SD and NS conditions. First, task performance levels may have differed in the two conditions. In fact, only two prior SD neuroimaging studies reported no observable behavioral deficit (Drummond et al., 2004; Portas et al., 1998), and Portas’ study found only thalamic responses to SD. Although our study is not immune from such a performance-based interpretation (some significant behavioral effects were observed, and we had relatively low power for detecting between-group differences in behavioral performance), we nevertheless found little evidence of consistent behavioral impairment following SD, as previously observed in complex tasks (Harrison and Horne, 2000).

Alternatively, subjects may have engaged in an effortful attempt to maintain performance under SD. Support here comes from our debriefing sessions, in which every subject described task performance as requiring more effort following SD. The simplest effort-based model would propose global upregulation of brain activity as effort increased, and would be consistent with SD > NS activity. The time course findings in Fig. 3, however, suggest that this simplest effort-based explanation is inadequate. In particular, the different time courses suggest that SD induced at least four qualitatively different regional responses: upregulation (3a), downregulation (3b), recruitment (3c) and dropout (3d). The NS > SD finding in BA 31 also challenges the simplest effort hypothesis. Alternative, regionally specific formulations of an effort hypothesis, however, remain as potentially viable explanatory candidates.

Third, it has been suggested that performance strategies may change with SD (Drummond and Brown, 2001). In fact, complex tasks can provide more strategic opportunities for enhancing (or maintaining) performance (Harrison and Horne, 2000), and SpaceDock provides several such options. We suggest, however, that our findings are not solely strategy related based on three observations: (1) we focused our analyses specifically on brain regions exhibiting consistent activity differences across subjects, and one would expect subjects naïve to the task to select different maintenance strategies, (2) we found no consistent behavioral changes across subjects suggesting consistent strategic alterations; and (3) at debriefing, our subjects revealed no awareness of changing their task performance strategy. Thus, while it is always difficult to eliminate the strategy change explanation, we believe such changes played a limited role in this study.

Finally, it is possible that certain regions of the brain are inherently more sensitive to SD. The fact that this and prior studies show substantial task-specificity of cerebral activity following SD argues partly against an inherent sensitivity hypothesis. However, relative measures such as fMRI (as opposed to, e.g. CMRglu) do not probe the brain uniformly – they are biased towards regions most involved in the task at hand. Thus, inherent, regional cerebral sensitivity to SD remains a viable explanation of available data.

In summary, while our data are inconsistent with several hypotheses alternative to cerebral compensation, some such explanations cannot be rejected. Looking at the qualitative range of observed time courses actually suggests a multiplicity of explanations: some regions display changes in the magnitude of otherwise significant activation (possibly reflecting cognitive reserve (Stern, 2002); others appear to exhibit compensatory recruitment/dropout, as discussed by Drummond et al. (2004). These along with inherent sensitivity, effort, and strategic hypotheses remain to be rigorously teased apart with additional SD studies on both simple and complex, open-ended tasks.


There are three principal limitations of our study. First, while behavioral performance was found to be reasonably equated across sleep conditions, the between-subject behavioral analyses suffered from low power for detection of significant behavioral effects. Moreover, some behavioral differences were observed in both within- and between-subject analyses. In particular, there was some evidence of an unresolved learning curve (our Order by Condition interaction). Given our counterbalanced orders, an unresolved learning curve would tend to reduce the observed performance differences. Regardless, behavioral differences across conditions cannot be excluded as potentially contributing to the neuroimaging results. Also, we did not address the role of expertise or automaticity in task performance, which could affect neurophysiological responses in the face of SD (Peres et al., 2000). Finally, and most limiting is the small subject pool, which precluded standard random effects analysis. A major change in scanner hardware (including field uniformity and receiver coils) as well as limited funds precluded addressing this low-N directly. Instead, we restricted our interpretation to only those regions exhibiting reliable activation changes across subjects. While appropriate, this provides only an initial measure of inferential support, warranting further investigation of complex tasks in SD to assess the robustness and the generality of these findings.

Summary and conclusions

Using a complex navigation task, we identified a coherent cerebral network exhibiting changes in activity associated with a single night of SD. Given a substantial lack of consistent performance differences between conditions, this work extends previous findings of adaptive cerebral responses to SD to the domain of complex, open-ended tasks. The pSTS proved the most prominently involved region, as found by Drummond et al. (2004) on a grammatical task. Additional work will be necessary to establish the repeatability of these findings across both subjects and tasks.


We acknowledge the assistance of Drs David Boas, Gary Jasdzewski, and Quan Zhang, as well as Jennifer Holmes. We acknowledge the support from the National Space Biomedical Research Institute through NASA Cooperative Agreement NCC 9–58, the NIH (NINDS: K25-NS046554; NCRR: P41-RR014075), and the Mental Illness and Neuroscience Discovery (MIND) Institute.