Trigeminal Nerve Tractography at 7T Ultra-High Field MRI: Validation and Description of a Technique

Diffusion tractography, derived from diffusion-weighted MRI (dMRI), provides noninvasive information about white matter architecture. However, its application to small structures such as trigeminal nerves (TGN) remains difficult due to susceptibility artifacts and limited spatial resolution. Ultra-high field strength MRI scanners, including 7-Tesla units, offer higher SNR and enhanced contrast and resolution compared with conventional clinical units. This may provide improved visualization and quantitative analysis of nerves. Since susceptibility artifacts are exacerbated at ultra-high field strengths, particularly in the skull base, signal loss has historically challenged TGN tracking. This study describes our methods for overcoming those obstacles, including tailored sequence design that permits robust tracking of the TGN in vivo.

7T-MRI data were acquired on seven trigeminal neuralgia (TN) patients. For the purpose of illustration, two TN patients with different etiologies are reviewed, one with classical TN and one with an epidermoid skull base tumor. Subjects underwent the MRI protocol listed in [Table 1], including high-angular-resolved, simultaneous multislice dMRI (68 directions). Localized shims and two different encoding directions were used to correct for brainstem B₀ artifact. Corrected dMRI images were then coregistered to T₁-weighted images.

Fiber orientation distributions were obtained from diffusion-weighted images corrected by spherical deconvolution. TGN trajectories were visualized within the brainstem using Second-Degree integration over Fiber Orientation Distribution (iFOD2) algorithm implemented in MRTRIX3 with carefully-placed regions of interest (ROIs). Three ROI seeds were placed in each nerve: anteriorly at Meckel’s cave entry, posteriorly at pons emergence, and within the ipsilateral pons. ROIs were drawn on coronal images, perpendicular to each nerve’s cisternal trajectory, thereby ensuring inclusion of the entire nerve cross-section. Pontine ROIs were placed at the top and bottom of the pons using overlaid T₁ and diffusion images. ROIs were drawn as axial lines on sagittal images and then filled through the entirety of the pons between the top and bottom limits on coronal images. Streamlines were required to pass through these ROIs along the nerve and down to the brainstem. Ultimate fiber tracking was reviewed by a neuroradiologist and a neurosurgeon for validation.

This approach represents evolution of trigeminal ROI placement following prior challenges in refining the tractography of the TGN and eliminating extraneous tracts, specifically from the brachium pontis. Placing three nerve seeds enables greater specificity and accuracy in detection of tracts than fewer seeds. In addition, our method that overcame difficulty defining tracts required enlargement of ROIs to encompass the entire nerve in the coronal plane to assure more accurate tract detection.

[Fig. 1] shows tracts visualized in classical and tumor-associated TN patients. TGN trajectory from Meckel’s cave and descending into the brainstem is highlighted. 7T-MRI holds promise for robust tracking and visualization of the trigeminal nerves. This may be exploited for quantitative analysis of microstructural nerve characteristics in relevant patient populations, including trigeminal neuralgia.

Table 1 Imaging scans parameters

Fig. 1 TGN tracts visualized in classical TN patient on (A) axial and (B) sagittal images and in tumor-TN patient on (C) axial and (D) sagittal images.

No conflict of interest has been declared by the author(s).