Internal states as a source of subject-dependent movement variability are represented by large-scale brain networks

• Published in: Nature communications Vol. 14; no. 1; p. 7837
• Main Authors: Breault, Macauley Smith, Sacré, Pierre, Fitzgerald, Zachary B., Gale, John T., Cullen, Kathleen E., González-Martínez, Jorge A., Sarma, Sridevi V.
• Format: Journal Article Web Resource
• Language: English
• Published: London Nature Publishing Group UK 29.11.2023; Nature Publishing Group; Springer Science and Business Media LLC; Nature Portfolio
• Subjects: 631/378/116; 631/378/116/2393; Behavior; Brain; Brain research; Computer engineering; Decision making; Electrodes; Engineering, computing & technology; Errors; Human subjects; Humanities and Social Sciences; Ingénierie, informatique & technologie; Motor task performance; multidisciplinary; Networks; Perturbation; Science; Science (multidisciplinary); State space models; Substrates
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-023-43257-4

Humans’ ability to adapt and learn relies on reflecting on past performance. These experiences form latent representations called internal states that induce movement variability that improves how we interact with our environment. Our study uncovered temporal dynamics and neural substrates of two states from ten subjects implanted with intracranial depth electrodes while they performed a goal-directed motor task with physical perturbations. We identified two internal states using state-space models: one tracking past errors and the other past perturbations. These states influenced reaction times and speed errors, revealing how subjects strategize from trial history. Using local field potentials from over 100 brain regions, we found large-scale brain networks such as the dorsal attention and default mode network modulate visuospatial attention based on recent performance and environmental feedback. Notably, these networks were more prominent in higher-performing subjects, emphasizing their role in improving motor performance by regulating movement variability through internal states. How internal states such as confidence and motivation influence motor performance remains unclear. Here, the authors explore brain networks associated with these internal states, finding that the Dorsal Attention Network encodes error states and the Default Network reflects perceived uncertainty.