Fatigue Leads to Dynamic Shift in Fronto-parietal Sustained Attention Network

Abstract

Recently, the gamma band ( γ; 70-100 Hz) has been implicated in sustained attention decay across a vigil consistent with computational models of fatigue. Frontal γ indexing centrally controlled sustained attention and parietal γ linked to gated sensory processes declined across a 10-minute vigilance task, a pattern observed for faster but not slower performers. The anatomical distribution of γ activity indicates neural communication, or connectivity, within the fronto-parietal network. We used Granger Prediction to evaluate fatigue effects on network γ connectivity. Results showed stronger directional connectivity for frontal→parietal versus parietal→frontal over time, indicating that top-down control of attention largely remained intact. However, parietal→frontal early γ connectivity increased with time, suggesting a network shift to enhanced sensory-directed processes after only 8 minutes. This pattern of connectivity was mirrored by fast but not slower performers. Our findings provide new directions for computational accounts of fatigue mechanisms and highlight the importance of individual differences.

Enhanced Bidirectional Gamma-band Granger Prediction in Fast Compared to Slow Performers in a 3-stimulus Oddball Task

“…Gamma-band Granger predictions were calculated at the midline in the frontal to parietal (sensor Fz → sensor Pz) and parietal to frontal (Pz → Fz) directions in two time/frequency windows, early (50 - 175 ms after stimulus onset, 70 - 85 Hz), and late (225 - 325 ms, 70 - 85 Hz). Our GP processing strategies and mathematics are fully described elsewhere (Borghetti et al, 2022). Time window and frequency range selections were based on the literature (e.g., Kim et al, 2017) and inspection of aggregate time-frequency GP plots.…”

“…A hallmark of human information processing, central cognition governs controlled and goal-oriented processes – the ability to attend to the task at hand, select contextually relevant stimuli, and adapt flexibly to the environment (Botnivick & Braver, 2015). In our previous work, we demonstrated high frequency gamma (γ) oscillations (~70-100 Hz) index controlled processes in a simple reaction time, or SRT, task (Borghetti et al, 2021; Borghetti et al, 2022; Curley et al, 2022). SRT tasks advantageously constrain neural signals to a single controlled process (i.e., sustained attention) and our studies revealed the effect of fatigue on reactive control –defined as processes transiently recruited on an as-needed basis such as stimulus detection (Braver, 2012; Cooper et al, 2016).…”

Gamma Power as an Index of Sustained Attention in Simulated Vigilance Tasks

“…We provide a quick overview of the behavioral and electrophysiology methods below; however, further details can be found in Borghetti et al (2021) and Borghetti et al (2022).…”

“…Further details on the sampling and cleaning procedures can be found in research by Borghetti et al. (2021, 2022).…”

“…For the gamma spectral analysis, we assayed power in the 70-100 Hz frequency band for frontal (Fz) and parietal (Pz) cortical regions. Further details on the sampling and cleaning procedures can be found in research by Borghetti et al (2021Borghetti et al ( , 2022.…”