Extraction of practice-dependent and practice-independent finger movement patterns.

Extensive motor practice can reorganize movements of a redundant number of degrees of freedom (DOFs). Using principal component (PC) analysis, the present study characterized the movement reorganization of the hand that possesses a large number of DOFs during a course of practice. Five musically naïve individuals practiced to play a short sequence of melody with the left hand for four successive days, and their hand kinematics was measured using a motion capture system. The PC analysis of the hand joint kinematics identified two distinct patterns of movement, which accounted for more than 80% of the total variance of movements. The second PC but not the first PC changed through practice. A correlation analysis demonstrated that the PC sensitive to the practice was characterized by coupled movements across fingers in the same direction. A regression analysis identified a decrease in the contribution of this PC to the hand movement organization through practice, which indicates a reduction of the movement covariation across fingers and thus an enhancement of the individuated finger movements. The results implicate potential of PC analysis to extract practice-invariant and practice-dependent movement patterns distinctively in complex hand motor behaviors.