Abstract
The emergence of distributed patterns of neural activity supporting brain functions and behavior can be understood by study of the brain’s low-dimensional topology. Functional neuroimaging demonstrates that brain activity linked to adaptive behavior is constrained to low-dimensional manifolds. In human participants, we tested whether these low-dimensional constraints preserve working memory performance following local neuronal perturbations. We combined multi-session functional magnetic resonance imaging, non-invasive transcranial magnetic stimulation (TMS), and methods translated from the fields of complex systems and computational biology to assess the functional link between changes in local neural activity and the reshaping of task-related low dimensional trajectories of brain activity. We show that specific reconfigurations of low-dimensional trajectories of brain activity sustain effective working memory performance following TMS manipulation of local activity on, but not off, the space traversed by these trajectories. We highlight an association between the multi-scale changes in brain activity underpinning cognitive function.
Highlights
The emergence of distributed patterns of neural activity supporting brain functions and behavior can be understood by study of the brain’s low-dimensional topology
To what degree, the reshaping of low-dimensional trajectories is causally affected by local neural perturbations remains unknown. Addressing this question is critical to bridge the knowledge gap on the causal relation between local and system-wide brain activity patterns supporting cognitive functions[10]. We addressed this knowledge gap by combining within-subject multi-session neuroimaging, a validated working memory task (n-back), a non-invasive brain stimulation approach known to induce changes to local neural activity, and methods translated from the fields of complex systems and computational biology[12]
We show that targeted perturbations of neural activity on a task-relevant brain region rather than a region not engaged by the task (S1) caused a specific reshaping of low-dimensional trajectories of brain activity that relates to changes in task performance
Summary
The emergence of distributed patterns of neural activity supporting brain functions and behavior can be understood by study of the brain’s low-dimensional topology. We defined session (baseline, S1, iPS), task (low versus high cognitive load), and behaviorally specific (correct versus incorrect performances) low-dimensional trajectories using data reduction and a state-of-the-art embedding technique (potential of heat diffusion for affinity-based transition embedding method, PHATE12).
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call