Artikel
Diffusion tensor imaging based language fiber tract navigation in glioma surgery
Suche in Medline nach
Autoren
Veröffentlicht: | 20. Mai 2009 |
---|
Gliederung
Text
Objective: To prevent speech deficits in glioma surgery by integrating the course of fibers connecting language related brain areas into a microscope-based navigation setup.
Methods: Diffusion tensor imaging based fiber tracking was performed applying diffusion weighted echo planar imaging on a 1.5T MR. Broca and Wernicke areas were identified by functional MRI and used as seed regions for tracking. A tensor-deflection algorithm was applied to reconstruct language related fiber tracts, which were rigidly co-registered with 3-D anatomical data for navigation integration. Additionally, fiber tracts reconstructed applying probabilistic tracking (A*-algorithm) were used for validation. Envelopes wrapping the reconstructed fiber bundles were generated automatically and visualized in the surgical field using operating microscope head-up displays.
Results: In 36 glioma patients (WHO grade I: 2, II: 6, III: 14, IV: 14) language fiber tracts, motor, and sensory language areas were visualized successfully into the surgical field. Navigation accuracy documented as target registration error was 1.7±0.7 mm. The differences between the 3-D objects generated by both tracking strategies were less than 5 mm. The image distortion of the B0 diffusion images impeded registration for navigation integration with less than 2 mm in the areas of interest.
In 9 patients (25%) (among which were 5 with preoperative speech deficits) we observed a speech deterioration which resolved only partially in 2 (5.6%) (both had no preoperative deficits) and completely reached the preoperative status in 7 (19.4%).
Conclusions: The course of language fiber tracts can be visualized reliably within the operative field, allowing removal of gliomas adjacent to language areas with low morbidity.