Targeting Lower Limb, Upper Limb, and Face Representation in the Primary Motor Cortex for the Practice of Neuronavigated Transcranial Magnetic Stimulation

Abstract

ObjectiveThe primary motor cortex (M1) is a usual target for therapeutic application of repetitive transcranial magnetic stimulation (rTMS), especially the region of hand motor representation. However, other M1 regions can be considered as potential rTMS targets, such as the region of lower limb or face representation. In this study, we assessed the localization of all these regions on magnetic resonance imaging (MRI) with the aim of defining three standardized M1 targets for the practice of neuronavigated rTMS. Materials and MethodsA pointing task of these targets was performed by three rTMS experts on 44 healthy brain MRI data to assess interrater reliability (including the calculation of intraclass correlation coefficients [ICCs] and coefficients of variation [CoVs] and the construction of Bland-Altman plots). In addition, two “standard” brain MRI data were randomly interspersed with the other MRI data to assess intrarater reliability. A barycenter was calculated for each target (with x-y-z coordinates provided in normalized brain coordinate systems), in addition to the geodesic distance between the scalp projection of the barycenters of these different targets. ResultsIntrarater and interrater agreement was good, according to ICCs, CoVs, or Bland-Altman plots, although interrater variability was greater for anteroposterior (y) and craniocaudal (z) coordinates, especially for the face target. The scalp projection of the barycenters between the different cortical targets ranged from 32.4 to 35.5 mm for either the lower-limb-to-upper-limb target distance or the upper-limb-to-face target distance. ConclusionsThis work clearly delineates three different targets for the application of motor cortex rTMS that correspond to lower limb, upper limb, and face motor representations. These three targets are sufficiently spaced to consider that their stimulation can act on distinct neural networks.