Neuroergonomics; sleep deprivation; fmri; granger causality; effective connectivity


The comprehensive effect of sleep deprivation on biological and behavioral functions largely remains unknown. There is evidence to support that human sleep must be of sufficient duration and physiological continuity to ensure neurocognitive performance while we are waking. Insufficient sleep would lead to high risk of human-error related to accidents, injuries or even fatal outcomes. However, in modern society, more and more people suffer from sleep deprivation because of the increasing social, academic or occupational demand. It is important to study the effect of sleep deprivation, not only on task performance, but also on neurocognitive functions. Recent research that has explored brain effective connectivity has demonstrated the directed inference interaction among pairs of brain areas, which may bring important insight to understand how brain works to support neurocognitive function. This research aimed to identify the brain effective connectivity pattern associated with changes of a task performance, response time, following sleep deprivation. Experiments were conducted by colleagues at Neuroergonomics Department at Jagiellonian University, Krakow, Poland. Ten healthy young women, with an average age of 23-year-old, performed visual spatial sustained-attention tasks under two conditions: (1) the rest-wakeful (RW) condition, where participants had their usual sleep and (2) the sleep-deprived (SD) condition, where participants had 3 hours less sleep than their usual sleep, for 7 nights (amounting to 21 h of sleep debt). Measures included eye tracking performance and functional magnetic resonance imaging (fMRI). In each condition, each subject*s eye-position was monitored through 13 sessions, each with 46 trials, while fMRI data was recorded. There were two task performance measures, accuracy and response time. Accuracy measured the proportion of correct responses of all trials in each session. Response time measured the average amount of milliseconds until participants gazed at the target stimuli in each session. An experimental session could be treated as a short window. By splitting long trials of fMRI data into consecutive windows, Granger causality was applied based on short trials of fMRI data. This procedure helped to calculate pairwise causal influences with respect to time-varying property in brain causal interaction. Causal influence results were then averaged across sessions to create one matrix for each participant. This matrix was averaged within each condition to formulate a model of brain effective connectivity, which also served as a basis of comparison. In conclusion, significant effect of sleep deprivation was found on response time and brain effective connectivity. In addition, the change of brain effective connectivity after sleep deprivation was linked to the change of response time. First, an analysis of variance (ANOVA) showed significant difference for response time between the RW condition and the SD condition. No significant changes for accuracy were found. A paired t-test showed that response time was significantly shorter in sleep deprivation for the visual spatial sustained-attention task. Second, Granger causality analysis demonstrated a reduction of bidirectional connectivity and an increase of directed influences from low-level brain areas to high-level brain areas after sleep deprivation. This observation suggested that sleep deprivation provoked the effective connectivity engaged in salient stimuli processing, but inhibited the effective connectivity in biasing selection of attention on task and in maintaining self-awareness in day time. Furthermore, in the SD condition, attention at the visual spatial task seemed to be driven by a bottom-up modulation mechanism. Third, a relationship was found between brain effective connectivity with response time. Decreases of Granger causal influences in two directions, from medial frontal lobe to sub cortical gray nuclei and from medial parietal lobe to sub cortical gray nuclei, were associated with shorter response time in the SD condition. Additionally, an increase of Granger causal influence from medial parietal lobe to cerebellum was associated with longer response time in the SD condition.


If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at

Graduation Date





Karwowski, Waldemar


Doctor of Philosophy (Ph.D.)


College of Engineering and Computer Science


Industrial Engineering and Management Systems

Degree Program

Engineering and Computer Science








Release Date

June 2015

Length of Campus-only Access


Access Status

Doctoral Dissertation (Open Access)


Dissertations, Academic -- Engineering and Computer Science; Engineering and Computer Science -- Dissertations, Academic

Included in

Engineering Commons