P 13. Intersession reliability of cortical motor maps with navigated transcranial magnetic stimulation

Abstract

Navigated transcranial magnetic stimulation (nTMS) is a non-invasive highly effective tool for mapping individual muscle representations in the motor cortex Julkunen et al., 2009 , Picht et al., 2011 . Optical navigation and model based estimates of intracortical target sites and stimulation strength suggest that nTMS should provide spatially more precise cartography than using external landmarks for defining a motor ‘hot-spot’. The remaining outstanding challenges are (a) a high inter-and intra-subject variability of motor evoked potentials (MEP) Schmidt et al., 2009 , Brasil-Neto et al., 1992 , (b) the extent to which anatomical mapping relative to the underlying gyrification is preferable and (c) hemispheric functional asymmetries possibly reflecting functional “lateralization”. These factors are particularly relevant in a clinical context, where typically cartography will be used with single pulses per stimulation location. Further, the definition of a representation “area” is strongly confounded by the variability of “fringe neurons” or overlapping cortical representations (Schieber and Hibbard, 1993) . The aim was to study the extent to which single pulse cartography can assure reproducible results. In the present analysis we focus on the concept of the “hot-spot”, i.e. point of maximum MEP and Center of Gravity (CoG). 15 subjects were examined in two sessions by two different researchers. Stimulation was performed using a computer based navigated TMS system with a figure-of-eight-coil (eXimia, Nextim Ltd., Helsinki, Finland). Mapping of the FDI over both hemispheres was performed with perpendicular (fine) and changing (random) coil orientation with an average of 79 (±35.91) stimuli at 110% resting motor threshold. Margins were defined as scalp locations where a stimulus resulted in an MEP <50 μV. To assess intersession variability we calculated the intraclass correlation coefficient (ICC), which is considered to reflect high reliability when higher than 75% (Mortifee et al., 1994) . Stimulus location for the FDI-“hot-spot” in the dominant hemisphere showed a high intersession correlation (ICC: 81–87%). In the non-dominant hemisphere the ICC was lower (45–73%). There was no significant difference in ICC using intracortical instead of scalp stimulus location. Weighted “hot-spots” as expressed by the CoG showed a very high intersession reliability (ICC: 84–88%), the side of maximum MEP a rather poor intersession reliability (ICC: 9–72%). Perpendicular mapping was associated with significant higher test–retest correlation than random mapping (ICC for random mapping 22–38%). • Stimulus locations (“hot-spots”) in the dominant hemisphere are more precise than in the non-dominant hemisphere. • Intracortical locations are not superior to scalp location for “hot-spot”-measurements. • Weighted “hot-spots” are more precise than maximal MEPs for “hot-spot”-definition and inter-rater reproducible. • Perpendicular mapping is clearly spatially more accurate than random mapping.