Neural signatures of sustained attention across time scales and individuals

Neural signatures of sustained attention across time scales and individuals

by Monica Rosenberg

Life & Natural Systems Neuroscience PSANNeurosciencesPSAN5Cognitive and behavioural neuroscience
🎙 Monica Rosenberg 👥 6K 📅 April 6, 2026 ⏱ 31 min 👁 295 📄 original study 🧭 2026-08-16
Available in: English (current) Français

Keywords

sustained attentionfunctional connectivitypredictive modelingfMRIindividual differences

Summary

Monica Rosenberg presents her lab’s research on neural signatures of sustained attention. She begins by introducing the concept of attention dynamics across time scales, from subsecond fluctuations to developmental changes. The core of her talk focuses on using functional connectivity patterns from fMRI to predict individual differences in sustained attention. She describes a predictive modeling approach where whole-brain connectivity matrices are used to predict task performance, and she shows that these models generalize across datasets and populations, including children and adults. She also demonstrates that these connectivity signatures are sensitive to attentional states, as they change with pharmacological manipulations and fluctuate with moment-to-moment engagement. Importantly, she argues that the ability to predict traits may stem from capturing dynamic states during data collection. She supports this with a study using ‘annotated rest’ where participants reported their thoughts, and connectivity patterns predicted these reports. Finally, she discusses how these models can be used to understand interactions between attention and other cognitive processes like memory and surprise.

164 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talk provides valuable insights into the neural basis of sustained attention, offering a robust methodological framework (connectome-based predictive modeling) that has been validated across multiple studies and datasets. The argumentation is solid, building from initial findings to more complex questions about state and trait prediction. The speaker acknowledges limitations and alternative interpretations, which strengthens the credibility. However, some claims are forward-looking and based on preliminary evidence, which is appropriately flagged.

Scientific Rigor, Source Quality, Title Accuracy

The research is rigorous, with careful cross-validation and generalization to large independent datasets (ABCD, HCP). The speaker cites her own and collaborators’ work, and mentions methodological contributions from others (e.g., co-fluctuation method). The title accurately reflects the content. No external sources are provided in the description, but the talk references published work.

138 words

Title / Content Match

The title accurately reflects the content, which focuses on neural signatures of sustained attention across time scales and individuals.

Quality & Reliability

8/10

The talk presents original research from a leading lab, with robust methods (cross-validation, large datasets like ABCD and HCP), and transparent discussion of limitations. However, it is a conference presentation, not peer-reviewed, and some claims are forward-looking.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Contribution & Novelties

The talk presents a novel framework for understanding sustained attention by linking brain connectivity patterns to both trait-like individual differences and dynamic states. It introduces the idea that predictive models can capture states during data collection, which may explain trait prediction. This is a significant conceptual contribution.

Pour aller plus loin :

103 words

Radar Profile

The radar profile shows high scores in information quantity, quality, and reliability, with a slightly lower technical level, indicating a well-balanced presentation that is accessible yet scientifically rigorous.

Reliability 8/10