Artikel
Multimodal language mapping in patients with left-sided perisylvian brain lesions – A comparison of DCS, rTMS, AND fMRI
Suche in Medline nach
Autoren
-
Sebastian Ille
- Neurochirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, München, Deutschland
-
Nico Sollmann
- Neurochirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, München, Deutschland
-
Noriko Tanigawa
- Faculty of Linguistics, Philology, & Phonetics, University of Oxford, Oxford, United Kingdom
-
Bernhard Meyer
- Neurochirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, München, Deutschland
-
Florian Ringel
- Neurochirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, München, Deutschland
-
Sandro M. Krieg
- Neurochirurgische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, München, Deutschland
| Veröffentlicht: | 13. Mai 2014 |
|---|
Gliederung
Text
Objective: In this study we analyze the applicability of repetitive transcranial magnetic stimulation (rTMS) to patients with lesions in language-related areas of the brain. Thus we compared the results of rTMS to functional magnetic resonance imaging (fMRI), the current standard for preoperative language mappings and to direct cortical stimulation (DCS) during awake surgery. With these results we want to provide a protocol for rTMS language mapping and the interpretation of its raw data. Further, we make a proposal for the combination of the results of rTMS and fMRI.
Method: We performed language mapping in 39 patients with left-sided perisylvian lesions by DCS, rTMS, and fMRI. 13 rTMS mappings were performed with a picture to trigger interval (pti) of 0 ms. The language errors of the rTMS mappings were categorized and the error rates (ER = number of errors per number of stimulations) were calculated for each region of the cortical parcellation system (CPS). Subsequently the rTMS mappings were analyzed, using different error rate thresholds (ERT = ER, at which a CPS region was defined as language positive in terms of rTMS). Additionally we combined the results of fMRI and rTMS and compared them to DCS as well in a protocol of combined non-invasive mapping.
Results: The rTMS maps compared to DCS yielded an overall sensitivity of 97.2% (100% with 0 ms pti), a specificity of 12.6% (15.4% with 0 ms pti), a positive predictive value (PPV) of 32.1% (23.3% with 0 ms pti) and a negative predictive value (NPV) of 91.3% (100% with 0 ms pti). For the fMRI maps compared to DCS we obtained an overall sensitivity of 37.8%, a specificity of 85.2%, a PPV of 49.1% and a NPV of 78.4%. However, when, combining language maps of fMRI and rTMS as non-invasive techniques the comparison of protocol 1 (decrease rTMS false positive results by qualifying them by fMRI negative results) with DCS showed an overall sensitivity of 41.3%, a specificity of 81.9%, a PPV of 50.9%, and a NPV of 75.3%. Protocol 2 (decrease fMRI false negative results by qualifying them by rTMS positive results) revealed a sensitivity of 98.4%, a specificity of 12.3%, PPV 33.9%, and a NPV of 94.4%.
Conclusions: With this study we can show that in comparison to DCS, rTMS is a more accurate tool for preoperative language mapping than fMRI. Moreover, we can support previous data that rTMS mappings should be performed with a pti of 0 ms.
