Super-resolution Diffusion Tensor Imaging for Delineating the Facial Nerve in Patients with Vestibular Schwannoma.

Abstract

Objectives Predicting the course of cranial nerves (CNs) VII and VIII in the cerebellopontine angle on preoperative imaging for vestibular schwannoma (VS) may help guide surgical resection and reduce complications. Diffusion magnetic resonance imaging dMRI is commonly used for this purpose, but is limited by its resolution. We investigate the use of super-resolution reconstruction (SRR), where several different dMRIs are combined into one dataset. We hypothesize that SRR improves the visualization of the CN VII and VIII. Design Retrospective case review. Setting Tertiary referral center. SRR was performed on the basis of axial and parasagittal single-shot epiplanar diffusion tensor imaging on a 3.0-tesla MRI scanner. Participants Seventeen adult patients with suspected neoplasms of the lateral skull base. Main Outcome Measures We assessed separability of the two distinct nerves on fractional anisotropy (FA) maps, the tractography of the nerves through the cerebrospinal fluid (CSF), and FA in the CSF as a measure of noise. Results SRR increases separability of the CN VII and VIII (16/17 vs. 0/17, p = 0.008). Mean FA of CSF surrounding the nerves is significantly lower in SRRs (0.07 ± 0.02 vs. 0.13 ± 0.03 [axial images]/0.14 ± 0.05 [parasagittal images], p = 0.00003/ p = 0.00005). Combined scanning times (parasagittal and axial) used for SRR were shorter (8 minute 25 seconds) than a comparable high-resolution scan (15 minute 17 seconds). Conclusion SRR improves the resolution of CN VII and VIII. The technique can be readily applied in the clinical setting, improving surgical counseling and planning in patients with VS.