Article
Connectome analysis based on intraoperative awake mapping reveals new insights on language network organisation
Die Konnektom-Analyse basierend auf Sprach-Kartierung während Wach-Kraniotomien liefert neue Einblicke in die Organisation des Sprach-Netzwerks
Search Medline for
Authors
-
Sebastian Ille
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Haosu Zhang
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Nina Stassen
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Maximilian Schwendner
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Axel Schröder
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Bernhard Meyer
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
-
Sandro M. Krieg
- Klinikum rechts der Isar der TU München, Klinik und Poliklinik für Neurochirurgie, München, Deutschland
| Published: | May 25, 2022 |
|---|
Outline
Text
Objective: Direct cortical stimulation (DCS) and navigated transcranial magnetic stimulation (nrTMS) combined with diffusion tensor image (DTI) have been applied in daily neurosurgical practices to identify the language-related cerebral structures. However, previous investigations on their consistency only focused on cortical results but were unclear at subcortical and network levels. Network centrality is the essential parameter in network analysis to measure the importance of nodes in the network.
Methods: Cortical regions identified as language-relevant were recorded by nrTMS (preoperatively) and DCS (intraoperatively). For both modalities, whole-brain fiber tracking was constructed using Python library DIPY (Version 1.2.0, https://dipy.org), in which each voxel was used as the seed for the deterministic algorithm of fiber tracking under FAT at 0.1 together with FLT at 30mm, and angulation thresholding at 25o. Tracked fibers were registered to the MNI246 template to identify the number of fibers between every two regions for constructing the connectome of subcortical connections, based on which five kinds of centrality were analyzed using Python library NetworkX (Version 2.6.2; https://networkx.org) for each region.
Results: The current study enrolled 27 patients with brain tumors undergoing both nrTMS and DCS language mapping. The positive predictive value of cortical results in nrTMS language mapping reaches 63% compared to DCS mapping. They were consistent in tracking dorsal stream, including arcuate fasciculus and superior longitudinal fasciculus. Ventral stream tracts were better localized in nrTMS mapping. In addition, there is no apparent difference in network centralities between the two mapping methods, including Degree Centrality, Eigenvector Centrality, Closeness Centrality, Betweenness Centrality, and PageRank Centrality.
Conclusion: DCS and nrTMS are now proven to reveal the same subcortical network. The current study deepens the understanding of the organization of cerebral structures related to language function and provides a new angle to investigate the brain structures.
Notes: Authors Eric Suero-Molina and Kornelius Kerl have shared senior authorship.
