The neural basis of intermanual transfer in tool-use motor learning

To identify brain systems underlying IT in tool-use motor learning with resting-state fMRI.

Eighteen healthy right-handed adults volunteered for the study. For the behavioral task, we used a special pliers that open when gripped (i.e., opposite direction from normal). During a 16-minutes learning phase, participants used their left hand to either (1) pick up and move 144 small blocks with the pliers (training group) or (2) hold the pliers without moving them (control group). For each participant, tool-use skill for the left and right hands were assessed by measuring the time taken to move 20 small balls with the pliers (movement-time). Tool-use skills and resting-state fMRI were measured before and after motor learning. Resting-state functional connectivity (RSFC) was calculated between the right primary motor cortex (M1) and five regions of interest, i.e., the left M1, left and right supramarginal gyrus (SMG) and cerebellum.

Movement-time data and Z values for RSFC were each submitted to 2 x 2 analysis of variance with learning (before vs. after) and group (training vs. control) as effects of interest. For both hands, we found significant learning x group interaction, with the training group being faster in movement-time than the control group (p 0.05). In the training group, the magnitude of reduction in movement-time was correlated between the left- and right-hands (r = 0.90, p 0.001). In RSFC analyses, significant interaction was found between the right M1 and right SMG, with functional connectivity being greater for the training group relative to the control group (FDR-p 0.05). In the training group, this increase in RSFC correlated with the magnitude of reduction in movement-time for both the left and right hands (r = −0.85 p 0.05 and r = −0.72, p 0.05, respectively).

We found that motor learning with the left hand facilitated tool-use performance not only in the same hand but also in the right hand via IT, supporting the notion that effector-independent motor memory is generated for one hand and transferred to the opposite untrained hand (Kumar, 2020). This behavioral change was associated with increased functional coupling between the right M1 involved in final motor output and SMG involved in motor skill memory (Seidler, 2008). These structures thus may mediate tool-use skill learning in the right hand via interhemispheric memory transfer. 

Tool-use learning via IT in the untrained right hand is associated with M1-SMG functional coupling in the ipsilateral right hemisphere. Although additional data samples and more formal analysis are needed, these results provide a basis for developing novel training methods for tool-use learning with a paralyzed hand.