Intraoperative Evaluation of semiautomatic Localization of the facial nerve using Diffusion Tensor Imaging in patients with large vestibular schwannomas: A Pilot Study.

Background - Facial nerve damage remains a significant risk during vestibular schwannoma (VS) resection, with reported incidences varying widely (3–46%). Damage risk increases with tumor size. Digital tractography enables nerve reconstruction but typically involves manual procedures, resulting in subjective evaluations that limit reproducibility and validation. We introduce a robust, semi-automatic tractography methodology with reproducible region-of-interest (ROI) generation and present an initial validation using a novel quantitative 3D comparison approach in patients with a large VS. Objectives - To assess the accuracy of facial nerve reconstruction employing a semi-automatic ROI selection method in patients with vestibular schwannomas. Materials and Methods - We included 6 patients with an average tumor size of 28mm (95%CI 17-40, 100% left) that underwent translabyrith VS surgery. Each VS patient was scanned with the regular neuronavigation magnetic resonance imaging (MRI) protocol and a custom diffusion-MRI (dMRI) protocol before surgery. The facial nerve trajectory was reconstructed with a diffusion tensor imaging (DTI)-based tractography software package using semi-automatic ROI generation. We validated our reconstructions with the neuronavigation system for intraoperative point annotation along the course of the facial nerve. Results - Tracts could be reconstructed in all included patients. The median distance and angle between the points and closest reconstruction were 5.1mm ( IQR 3.5 – 7.6) and 38.5 degrees (IQR 2.7 – 79.8) respectively. Conclusion – We present a promising methodology for facial nerve reconstruction in patients with VSs. However, further optimization of the methodology is warranted before a proper clinical validation study can be performed.