Motor imagery training: Kinesthetic imagery strategy and inferior parietal fMRI activation

Abstract

Motor imagery (MI) is the mental simulation of action frequently used by professionals in different fields. However, with respect to performance, well-controlled functional imaging studies on MI training are sparse. We investigated changes in fMRI representation going along with performance changes of a finger sequence (error and velocity) after MI training in 48 healthy young volunteers. Before training, we tested the vividness of kinesthetic and visual imagery. During tests, participants were instructed to move or to imagine moving the fingers of the right hand in a specific order. During MI training, participants repeatedly imagined the sequence for 15 min. Imaging analysis was performed using a full-factorial design to assess brain changes due to imagery training. We also used regression analyses to identify those who profited from training (performance outcome and gain) with initial imagery scores (vividness) and fMRI activation magnitude during MI at pre-test (MI_pre ). After training, error rate decreased and velocity increased. We combined both parameters into a common performance index. FMRI activation in the left inferior parietal lobe (IPL) was associated with MI and increased over time. In addition, fMRI activation in the right IPL during MI_pre was associated with high initial kinesthetic vividness. High kinesthetic imagery vividness predicted a high performance after training. In contrast, occipital activation, associated with visual imagery strategies, showed a negative predictive value for performance. Our data echo the importance of high kinesthetic vividness for MI training outcome and consider IPL as a key area during MI and through MI training.

Keywords: fusiform gyrus; inferior parietal lobe; kinesthetic imagery; mental training; motor imagery; prediction of training gain; visual imagery.

Figure 1

Experimental procedure. After being given instructions about the experiment and the finger sequence tapping, and after filling in the MIQ‐R (visual (VI) and kinesthetic imagery (KI) score), the participant was placed in the MRI and performed a short prescan including the execution (EX) and mental imagery (MI) task. These scans and performance recorded during EX served as precondition. To control for possible extension movements during the MI task, the participant wore a virtual reality glove. During the 15‐min training, participants imagined the sequence for about 150 times. At post‐test, we assessed performance gain during EX task, and we performed a post‐training scan including the EX and MI task. The experimental session ended with a structural T1‐weighted image [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 2

Effect of MI training on the actual performance of the finger sequence. Number of errors (a) and velocity of tapping (b) decreased between pre‐ and post‐test (***p < .001)

Figure 3

(a) Effect of condition and training. The effect of condition (MI_pre − EX_pre) showed a left‐sided inferior parietal and medial temporal fMRI activation (orange). The effect of training (MI_post − MI_pre) showed a strong effect in the left inferior parietal lobe (blue). Statistical threshold: p < .05; FWE corrected for the whole volume. (b) Associations between fMRI activation during MI_pre and motor imagery score for kinesthetic (KI) and visual (VI) imagery. KI (orange) was positively associated with fMRI activation in right inferior parietal lobe during the MI_pre task. In contrast, VI (blue) was positively associated with fMRI activation in right visual striate. The statistical threshold has been adjusted to p < .05; FWE corrected per ROI. (c) Outcome prediction by fMRI activation during MI_pre. High performance after training (weighted velocity scores) could be predicted by high initial right fusiform activation (orange) during MI. Low performance after training could be predicted by high initial left occipital activation (blue) during MI_pre. Statistical threshold: p < .05; FWE corrected for the whole volume [Color figure can be viewed at http://wileyonlinelibrary.com]