Systematic Fluctuations of TMS-induced motor evoked potentials

Objectives: While latencies of transcranial magnetic stimulation (TMS) induced motor responses are relatively stable, measure of amplitude are notoriously variable (Kiers 1993) This leads to a wide range of normal values and limits its usefulness for single subjects or groups with minor differences (Magistris 1999; Wassermann 2002). The aim of this study was to investigate the correlation of motor evoked potentials (MEP) sweep-to-sweep variability with the variations of physical parameters of TMS. This should allow us to differentiate physiological and physical origins. Previous results show that trial variability is related to attention, motor imagery, expectancy and pre-stimulus contraction as well as stimulus intensity, electric field size and stimulation site. Here, we controlled with novel precision due to neuronavigation for the physical parameters translation, roll and yaw (figure-eight stimulation coil Magstim 200) as well as stimulation strength at and dislocation from a predefined target site (hotspot, fist digital interosseus muscle) with navigated brain stimulation. We argue that neurophysiological research should include precise measures of physical variations as these can be highly correlated with the sweep-to-sweep variability of MEPs.

Methods: Two hundred suprathreshold monophasic single pulse stimuli per subject (n=17) were applied on the dominant primary motor hand area with an intensity of 110% resting motor threshold. Evoked muscle potentials were recorded by belly-tendon electromyography of the contralateral first digital interosseous muscle. Stabilized mechanical fixation (Magic Arm, www.manfrotto.com) kept the root-mean-square error of physical variations below 2mm. The physical changes of stimulation parameters were captured with navigated TMS (nTMS, www.nexstim.com).

Results: Multiple linear regression found significant covariance between sweep-to-sweep variability and minor changes in physical parameters of TMS in all subjects. Physical variations account for 0–4%, in one subject for 80%, of the complete variability of MEPs.

Conclusions: Surprisingly, systematic physical variations account for a lesser part of sweep-to-sweep variability found in TMS-induced finger-twitches. This suggests that much of the observed variability is of neurophysiological origin. Possibly, after trial by trial correction of physical errors, this variability might offer a measure of cortical function in the unconscious or aphasic state (Ellaway 1998) and/or describe a noise-term to optimize performance (Harris & Wolpert 1998).