Leveraging the effector independent nature of motor imagery when it is paired with physical practice

The relationship between motor imagery and physical practice in skill acquisition remains a topic of ongoing research. While often considered analogous, recent work suggests a key difference: motor imagery primarily leads to effector-independent encoding of motor programs (global movement features, integration of perceptual information with movement goals, independent of specific effectors), while physical practice involves both effector-independent and effector-dependent encoding (mapping movement goals and parameters to the specific effector). This distinction is crucial because established skill acquisition frameworks highlight the necessity of both types of encoding for motor proficiency. The limited improvements in performance often seen with motor imagery alone might stem from this effector-independent nature. This study aimed to investigate this hypothesis by examining the impact of the order of motor imagery and physical practice on skill acquisition. It was hypothesized that if motor imagery is primarily effector-independent, then applying it before physical practice would lead to greater performance gains compared to applying it afterward.

Existing literature supports the idea that motor imagery and physical practice share neural representations, contributing to motor imagery's effectiveness in skill acquisition. However, recent studies challenge the assumption of functional equivalence, suggesting motor imagery results solely in effector-independent encoding, unlike physical practice which involves both effector-independent and -dependent encoding. This difference may explain why motor imagery alone yields less robust performance improvements. Studies examining inter-manual transfer after motor imagery training show improved performance in both trained and untrained effectors, contrasting with physical practice where performance is superior in the trained effector. While studies demonstrate the benefits of short-term motor imagery, research on multi-session training is scarce, especially concerning the interaction of motor imagery and physical practice. This study aimed to fill this gap by investigating the neural and behavioral consequences of combining these training modalities over a longer duration.

Thirty-eight right-handed participants were randomly assigned to one of three groups: MIP-PP, PP-MIP, and PP-PP. All participants completed 10 days of training on a dart-throwing task. MIP-PP involved 5 days of motor imagery followed by 5 days of physical practice, PP-MIP involved 5 days of physical practice followed by 5 days of motor imagery, and PP-PP involved 10 days of physical practice. Each training session consisted of 15 blocks of 6 dart throws. Participants in the MIP-PP and PP-MIP groups also underwent fMRI scans at three time points: pre-training, mid-training (after day 5), and post-training (after day 10). Dart-throwing performance was assessed using radial error (RE) and bivariate variable error (BVE), measuring accuracy and consistency respectively. Autocorrelation lag-1 (ACF1) was calculated to assess trial-by-trial corrections, reflecting effector-dependent encoding. Kinematic analysis (2D video) examined global kinematic variability and angular velocity. fMRI data was preprocessed and analyzed using FSL. Between- and within-group comparisons were performed to characterize functional brain changes during motor imagery.

After excluding two participants (one dropout, one excessive head motion during fMRI), 36 participants completed the study. No significant group differences in motor imagery ability were found. Linear mixed-effects models revealed significant interactions between group and time point for RE and BVE. The MIP-PP group showed greater improvements in RE at day 5 compared to the PP-MIP and PP-PP groups. While all groups reached similar skill levels by day 10, the MIP-PP group displayed superior overall improvements in RE and BVE compared to PP-MIP, though still inferior to PP-PP. Analysis of ACF1, an indicator of trial-by-trial corrections, showed a main effect of group but no significant time point or interaction effects. However, effect sizes indicated improvements in ACF1 primarily in the second half of training, regardless of imagery order. Kinematic analysis showed a significant decrease in global kinematic variability with training but no group differences. fMRI analyses revealed that training-related brain activation changes were only driven by physical practice and observed in the initial phase. Between-group comparisons showed that the MIP-PP group had increased ipsilateral cerebellum activation during pre-training, and PP-MIP group had increased activity in motor-related areas (SMA, ACC, cerebellum, precentral gyrus) during mid-training. Negative correlations were found between BVE and motor imagery-related activity in parietal and lingual gyri (mid-training) and frontal and occipital regions (post-training).

The findings challenge the long-held view of motor imagery being functionally equivalent to physical practice. The superiority of the MIP-PP group in overall performance and the patterns of brain activation support the hypothesis that motor imagery primarily facilitates effector-independent encoding. Improvements in consistency and global kinematics, along with the absence of significant changes in accuracy or trial-by-trial corrections (ACF1) following motor imagery, suggest that motor imagery primarily refines the global aspects of motor programs rather than the specific motor commands. The greater activation in motor-related areas in the PP-MIP group during mid-training, but not post-training, suggests that physical practice is essential for effector-dependent encoding. The negative correlation between BVE and activity in visual processing areas suggests that error correction during motor imagery may rely on these central executive functions rather than sensorimotor feedback.

This study demonstrates that motor imagery is not functionally equivalent to physical practice, primarily facilitating effector-independent encoding. Applying motor imagery before physical practice yielded better performance than the reverse order, although still less effective than physical practice alone. The findings highlight the importance of considering both effector-independent and -dependent encoding in motor learning research and suggest that optimal training strategies might involve strategically combining motor imagery and physical practice.

The study's limitations include the relatively small sample size and the potential for methodological limitations of fMRI to detect subtle changes in brain activity during the later stages of motor learning. The use of a block design in fMRI might have limited the detection of transient changes in brain activity. The relatively short training duration might not fully capture the long-term effects of motor imagery. Further research is needed to explore optimal training schedules and dosages of motor imagery combined with physical practice.