gms | German Medical Science

Objective: The preservation of important white matter tracts is extremely important to optimize the functional long-term outcome after brain tumour surgery. Depicting important fibres of the language network is very challenging. We here studied the use of navigated transcranial magnetic stimulation (nTMS) to depict the primary motor representation of the tongue (M1) for reconstruction of tracts connected to the motor-speech network.

Methods: 13 healthy volunteers were investigated by nTMS mapping of M1. The hotspot, i.e. the cortical site of the highest motor evoked potential (MEP) amplitude, of the tongue representation was localized. The 3D-coordinate of the respective voxel (site of max.calculated electrical-field strength) was visualized in a 3D view of the brain (FSL). The hotspot was then replaced at the nearest border between grey and white matter in order to allow for comparison with anatomical region-of-interest (ROI) seeding. Anatomical ROIs were placed at the same subcortical level by a blinded, experienced, second investigator in the middle of the lateral third of the precentral gyrus. Both ROIs were enlarged by r= 5 mm and served as origins for probabilistic fibre tracking (FSL). The reconstructed fibres were segmented according to anatomical knowledge. Tracts were visualized with MRIcron and 3D Slicer. The anatomy- vs. nTMS-based tractography results were compared regarding the reconstruction of the following tracts using McNemar's Chi-squared test for paired binomial data: short anterior / short posterior / long segment of the arcuate fasciculus (saAF, spAF, lAF) and frontal Aslant tract (frAT).

Results: Using M1 of the tongue as a single seeding-ROI for tractography, was a highly robust technique for reconstruction of at least one segment of the AF (TMS:N=13/13; anatomical:N=11/13; n.s.). In comparison of anatomical vs. nTMS-based seeding-ROI techniques regarding the different segments/ tracts, only the saAF was reconstructed significantly more often when using the nTMS-ROI (nTMS: 85%; anatomical: 31%) whereas no difference was observed for the spAF (nTMS: 62% vs. anatomical: 85%) and the lAF (nTMS: 92% vs. anatomical: 85%). Apart from the AF, also the frAT, which seems to play a crucial role for speech production, was reconstructed in most cases using the tongue-M1-seed for tractography (both 85%). Preliminary results regarding diffusion metrics and numbers of aberrant fibres point towards the use of nTMS for seeding-ROI determination being advantageous compared to anatomical ROI-seeding.

Conclusion: Using the M1 tongue core region as a single seeding-ROI in diffusion tractography allows to depict motor-speech associated fibres like the anterior segment of the AF. Compared to anatomical ROI-seeding, this approach may offer a higher accuracy and thus avoid erroneous fibre reconstruction, particularly in areas of unfavourable signal-to-noise-ratio.