Abstract WMP48: Targeting The Action Selection Network During Functional Arm Training After Stroke: A Feasibility Study

Abstract

Introduction: Interventions aimed at improving arm function after stroke primarily focus on movement execution. Approaches that target a specific motor planning process during training have been suggested but not systematically studied. Action selection (AS), or selection of a motor response, is an important planning process with well described behavioral (planning time) and neural (frontal-parietal network) correlates. The purpose of this study was to determine the feasibility of adding AS demands to task-oriented arm training in individuals post-stroke. Methods: Thirty individuals with chronic stroke (59.3±11.0 years of age; 34.9±35.6 months post-stroke) and mild to moderate motor impairment (Fugl-Meyer motor score: 41.4±14.7) completed 15 sessions of task-oriented arm training that included AS demands through abstract visual cues; cues dictated movement task or movement direction/distance/height. AS difficulty was progressed throughout training. Before and after training, individuals completed a joystick-based AS task during functional MRI. Results: Despite the added planning demand, movement repetitions were high (averaged 190.1±25.4 per session). AS accuracy (correct response selection) averaged more than 90% across sessions, even as AS difficulty increased. AS response times and activation in the AS neural network decreased from pre- to post-training, especially in premotor and prefrontal cortices. Gains were seen in upper extremity function (Action Research Arm Test: Pre: 38.0±16.3; Post: 41.4±15.3) and hand-related quality of life (Stroke Impact Scale Hand Domain: Pre: 47.2±23.7; Post: 61.8±24.1). Conclusions: Overall, it was feasible to add AS demands to functional arm training after stroke. Targeted engagement of the AS neural network through the addition of AS demands to arm training may challenge this component of the motor network in a manner that improves motor planning performance, upper extremity function, and neural resource utilization.