Attentional brain rhythms during prolonged cognitive activity

Abstract

As routine and lower demand cognitive tasks are taken over by automated assistive systems, human operators are increasingly required to sustain cognitive demand over long periods of time. This has been reported to have long term adverse effects on cardiovascular and mental health. However, it remains unclear whether prolonged cognitive activity results in a monotonic decrease in the efficiency of the recruited brain processes, or whether the brain is able to sustain functions over time spans of one hour and more. Here, we show that during working sessions of one hour or more, contrary to the prediction of a monotonic decline, behavioral performance in both humans and non-human primates consistently fluctuates between periods of optimal and suboptimal performance at a very slow rhythm of circa 5 cycles per hour. These fluctuations are observed in both high attentional (in non-human primates) and low attentional (in humans) demand conditions. They coincide with fluctuations in pupil diameter, indicating underlying changes in arousal and information-processing load. Accordingly, we show that these rhythmic behavioral fluctuations correlate, at the neurophysiological level, with fluctuations in the informational attention orientation and perception processing capacity of prefrontal neuronal populations. We further identify specific markers of these fluctuations in LFP power, LFP coherence and spike-field coherence, pointing towards long-range rhythmic modulatory inputs to the prefrontal cortex rather than a local prefrontal origin. These results shed light on the resilience of brain mechanisms to sustained effort and have direct implications on how to optimize high cognitive demand working and learning environments.

Revisiting Persistent Neuronal Activity During Covert Spatial Attention

“…In this context, previous studies have shown that when attention is actively sustained in time, such as in the context of long-lasting cognitive demands, and the performance seems to decrease (Proctor et al, 1996;Lockley et al, 2004;Bonnefond et al, 2010;Virtanen and Kivimäki, 2018). A recent work by Gaillard et al (2021) suggests that this might not always be the case. Indeed, they report that behavioral performance in a visual attentional task fluctuates by up to 10% at an ultra-slow rhythm of 4-7 cycles per hour (every 9-15 min), coinciding with phaselocked rhythmic fluctuations in the accuracy of visual and spatial attention information in the PFC.…”

Non-invasive real-time access to spatial attention information from 3T fMRI BOLD signals

“…Accordingly, training a classifier on trials for which attention was initially decoded closest to expected target position enhances the correlation between decoding accuracy and behavior (De Sousa et al, 2021). In addition, tracking the attentional spotlight at a high temporal resolution reveals that it is not stable but rather, it moves around at a frequency of 8 to 12 Hz (Gaillard et al, 2021). Extending these observations to the current non-invasive decoding of attention, we would like to suggest that the specific structure of the confusion matrices described in figures 3B and 4B, does not reflect decoding performance on noisy signal.…”

“…Extending these observations to the current non-invasive decoding of attention, we would like to suggest that the specific structure of the confusion matrices described in figures 3B and 4B, does not reflect decoding performance on noisy signal. Rather, it captures the dynamic nature of spatial attention, such that subjects, while trying hard to focus their attentional spotlight at the cued location cannot prevent it to explore space at close by locations (Gaillard et al, 2021). All this taken together strongly suggests that due to the dynamic nature of spatial attention function, the quest for ever improved decoding accuracy is an unreachable Grail.…”

Distractibility and impulsivity neural states are distinct from attention states and modulate the implementation of attention

“…Prior studies have shown that task-independent neural states might influence different aspects on how information is encoded in a neural population and, hence, its impact in behavior (Astrand et al, 2016;Cowley et al, 2020;Gaillard et al, 2021). For example, the level of shared noise correlation between neurons predicts subjects' behavior (Astrand et al, 2016;Ben Hadj Hassen et al, 2019;Nogueira et al, 2020).…”

“…Attention plays a critical role in all of these situations, implementing the selection of the sensory cues that are relevant to our ongoing purposes (Ibos et al, 2013;Moore and Armstrong, 2003;Wardak, 2006;Wardak et al, 2004Wardak et al, , 2002 while suppressing the irrelevant information the response to which has to be suppressed (Di . However, other factors independent from the goals of the task are also expected to interfere, either enhancing or degrading behavioral performance, such as fatigue (Marcora et al, 2009;Rosa et al, 2020), motivation (Brown and Bray, 2019;Di Bello et al, 2021), the degree of liberal or conservative biases in response performance (Cowley et al, 2020) or intrinsic fluctuations of information coding of cognitive processes (Gaillard et al, 2021). All of these factors arguably define "internal states" that highly influence the perceptual outcomes under similar sensory conditions.…”