Resting State Connectivity Is Modulated by Motor Learning in Individuals After Stroke

 

I have no understanding of how this helps recovery.  In my case since most of my premotor cortex is dead there can be no transfer from prefrontal to premotor.  And since I'm not healthy it wouldn't transfer anyways. When I follow the higher cognitive load reference²⁴ I find nothing explaining that at all. I was hoping for some explanation of the mental fatigue post stroke.

Resting State Connectivity Is Modulated by Motor Learning in Individuals After Stroke

Show all authors

Sarah N. Kraeutner, PhD, Cristina Rubino, MSc, Shie Rinat, MSc, ...

First Published April 7, 2021 Research Article 

https://doi.org/10.1177/15459683211006713

Article information

Abstract

Objective

Activity patterns across brain regions that can be characterized at rest (ie, resting-state functional connectivity [rsFC]) are disrupted after stroke and linked to impairments in motor function. While changes in rsFC are associated with motor recovery, it is not clear how rsFC is modulated by skilled motor practice used to promote recovery. The current study examined how rsFC is modulated by skilled motor practice after stroke and how changes in rsFC are linked to motor learning.

Methods

Two groups of participants (individuals with stroke and age-matched controls) engaged in 4 weeks of skilled motor practice of a complex, gamified reaching task. Clinical assessments of motor function and impairment, and brain activity (via functional magnetic resonance imaging) were obtained before and after training.

Results

While no differences in rsFC were observed in the control group, increased connectivity was observed in the sensorimotor network, linked to learning in the stroke group. Relative to healthy controls, a decrease in network efficiency was observed in the stroke group following training.

Conclusions

Findings indicate that rsFC patterns related to learning observed after stroke reflect a shift toward a compensatory network configuration characterized by decreased network efficiency.

Keywords motor learning, stroke, rs-fMRI, graph theory, functional connectivity

Introduction

Damage resulting from stroke disrupts cortical networks and patterns of synchronized brain activity between disparate brain regions (termed functional connectivity).^1-3 Synchronized patterns of brain activity can be characterized across the brain at rest (ie, resting-state functional connectivity [rsFC]) and their relationships represented as coherence. These patterns characterize functional reorganization of the brain after stroke and are reliable measure that characterize neural changes across the stages of recovery.⁴ While altered rsFC is associated with motor recovery (ie, improvements in function characterized by clinical assessments),⁵ it is not clear how rsFC is modulated by skilled motor practice after stroke (ie, behavioral improvements associated with a specific motor task). Even though rsFC does not rely on task performance, there is evidence showing that active networks mapped with rsFC overlap with regions involved in task performance.^6-8 As rsFC does not rely on participant effort or compliance, it may be used to characterize neural changes that accompany motor impairment poststroke. Typically, in individuals with stroke, rsFC is disrupted in the sensorimotor network relative to healthy individuals.^9,10 Increases in rsFC in both the sensorimotor network and between regions implicated in cognitive processes (ie, working memory) have been observed as motor recovery is achieved.^9,11,12 For instance, poorly recovered individuals showed decreased connectivity within the sensorimotor network, while no differences in connectivity were observed between individuals who were well-recovered and healthy controls.⁹ Yet a typical pattern of connectivity is not necessarily restored during recovery after stroke. Even in well-recovered individuals, relative to healthy controls, reduced connectivity persists between brain regions associated with cognitive processes.^9,13 To date, changes in rsFC have largely characterized functional reorganization that occurs in association with recovery from stroke.^11,12,14,15 It remains unclear whether or not skilled motor practice drives changes in rsFC patterns.

Importantly, rsFC is thought to reflect the processing of information gained during skilled motor practice associated with motor consolidation and learning.^16,17 Short-term changes in rsFC in areas previously shown to be critical to planning and executing visually guided movement^18,19 including a network of frontal, posterior parietal, and cerebellar regions are associated with learning a visuomotor task.^16,17,20,21 However, as behavioral change associated with task-specific learning plateaus, limited long-term changes in rsFC in healthy individuals are noted.²¹ While learning (and relearning) motor skills are critical to promoting functional recovery, we know little about the alterations in processes underlying motor learning after stroke. Thus, rsFC can be employed to characterize change in consolidation of motor memories and learning that result from functional reorganization after stroke.

Specifically, functional magnetic resonance imaging (fMRI) shows that healthy individuals shift brain activity from the prefrontal regions early in skilled motor practice to premotor cortical activation after learning occurs.^16,22 This shift is not observed after stroke.²³ The persistent and greater recruitment of frontal-parietal regions during motor tasks may reflect higher cognitive load during skilled motor practice after stroke.²⁴ It also may be related to an overall decrease in network efficiency after stroke,^25,26 that represents a lower overall capacity to transmit information and indicates that a compensatory network (ie, not restored to a neurotypical pattern of functioning) underlies motor processes.²³ Taken together, alterations in consolidation and learning processes may arise after stroke, reflected by decreased network efficiency and greater reliance on cognitive processes during skilled motor practice. To test this idea, we probed (long-term) changes in rsFC induced by skilled motor practice to examine how brain reorganization supports learning after stroke.

The primary aim of the current study was to examine how rsFC is modulated by skilled motor practice after stroke. Furthermore, we sought explore how changes in rsFC are linked to motor learning. To address our objectives, we employed a between-group design whereby 2 groups of participants (individuals with stroke and age-matched controls) engaged in 4 weeks of skilled motor practice of a complex, gamified reaching task, that was designed to prevent early plateaus in performance. Clinical assessments of motor function and impairment, and brain activity were obtained before and after training.

We expected that rsFC would be differentially modulated from pre- to posttraining between groups. Because past work showed that individuals with stroke rely on prefrontal regions during skilled motor practice,^23,27,28 and that greater recovery is linked to increased functional connectivity of frontal regions implicated in working memory,^9,24 we expected to observe increased connectivity within the sensorimotor network, and between the sensorimotor network and prefrontal areas. In exploring the association between changes in rsFC and motor learning, we hypothesized that (1) improvements in motor behavior associated with task-specific learning would be related to decreased connectivity between regions implicated in working memory in individuals with stroke and (2) healthy controls would show minimal connectivity changes. Finally, we predicted that changes in network efficiency induced by skilled motor practice would occur differentially after stroke relative to healthy controls. Specifically, we predicted that healthy controls would show enhanced network efficiency that would reflect their increased capacity to transmit information. In contrast, we expected that individuals with stroke would show decreases in network efficiency reflecting a shift toward a compensatory network configuration to support learning.