Fig. 3: Finger tapping movement profiles are predicted by SM1 evoked cortical responses in a contact-dependent manner.

a Conceptual figure denoting the statistical model probed in the current study to derive data-driven accelerometer-based movement profiles of finger tapping. The factors contributing to finger tapping movement profiles were obtained from an EFA including acceleration magnitude, acceleration variability (i.e., coefficient of variation) and reverse-coded tapping frequency with lower values of finger tapping movement profiles reflective of smoother movements during finger tapping blocks. b LME of finger tapping movement profiles scores as a function of SM1 response amplitude (continuous variable), contact orientation (factor with 5 levels) and their interaction were conducted controlling for acquisition order and total electrical energy delivered. There was a significant SM1 response by contact orientation interaction such that when clinically-effective contact orientations were applied, larger SM1 response amplitudes were predictive of smoother finger tapping movement profiles. 95% confidence intervals are displayed in gray for each regression line.