Fine-grained upper limb force control relies on distributed functional connectivity during motor planning: a scalp EEG study

Motor behavior consists of two distinct stages, i.e., planning and execution. However, the underlying brain network patterns that distinguish these two stages remain insufficiently understood. Moreover, the modulation of these network dynamics by different levels of motor force, such as those in daily grasping tasks, has been given limited consideration. In this study, we simultaneously recorded electroencephalogram (EEG) and electromyography (EMG) signals while participants performed upper limb grasping at different percentages of maximum voluntary contraction (MVC). Our results initially indicated that high-force conditions (100% MVC) generated more pronounced movement-related cortical potentials (MRCPs) compared to low-force conditions (20% MVC). Functional connectivity (FC) analysis further demonstrated that motor planning was associated with a more extensive network pattern, especially across the frontal, parietal, and temporal lobes, than motor execution. Additionally, during the planning stage, 20% MVC induced stronger long-range FC, particularly between fronto-occipital and fronto-parietal regions. Conversely, during execution, 100% MVC resulted in stronger FC mainly between frontal and central regions. Based on these network disparities, the classification accuracy for motor stages was 81.73% ± 11.25% under the 20% MVC condition and 84.65% ± 22.48% under the 100% MVC condition. Regarding force strengths, the classification accuracy was 75.58% ± 14.68% during planning and 90.39% ± 7.01% during execution. These results suggest that motor planning necessitates the recruitment of additional cognitive networks to support decision-making processes, especially for fine-grained motor tasks like small force control, whereas motor execution depends more significantly on sensorimotor networks to maintain movement performance. This functional differentiation highlights a hierarchical organization of motor control at the brain network level, thereby deepening our understanding of how distributed neural circuits support distinct stages of motor behavior.