Children with Neurofibromatosis type 1 (NF1) are predisposed to optic pathway gliomas (OPGs) that can cause uncorrectable vision loss. In children with NF1-associated OPGs, DTI measures of the optic radiations [fractional anisotropy (FA), radial diffusivity (RD), mean diffusivity (MD)] have been associated with visual acuity and may be an important tool for the evaluation of these tumors.[1] However, the optic radiation is a thalamo-cortical tract with start and end points that are more challenging to depict than those of cortico-cortico tracts, and diffusion MR tractography that relies on hand-drawn regions of interest requires user expertise that limits consistency between sites and studies. We created an automated method to identify white matter tracts in the optic radiations and compared this method to tractography using hand-drawn regions of interest as previously reported.[1] As a test of validity, we evaluated the association between each method’s DTI measures and visual acuity to determine whether automated tractography was equivalent to hand-drawn methods.

In 50 children with NF1 and OPG, we measured the association between FA, RD and MD of the optic radiations and visual acuity measured within 3 months of imaging. Visual acuity was measured by a pediatric ophthalmologist and converted to the logarithm of the minimal angle of resolution (logMAR) to provide a linear scale. A vision deficit in the worse eye of ≥0.2 logMAR from age norms was considered abnormal (approximately a 2-line drop on the Snellen chart). Multivariable linear regression additionally adjusted for age, tumor location, and DTI parameters. All MR examinations were performed at 3T on either a Trio, Skyra or Verio (Siemens; Erlangen, Germany). Diffusion MR was acquired with an echo planar pulse sequence with 128 x 128 matrix, in-plane voxel size of 2 x 2mm, diffusion weighting of b=1000 s/mm2, and full brain coverage with no gap between slices. Examinations were acquired with 30 gradient directions and 2mm slice thickness, except for 13 examinations that were acquired with 20 diffusion gradient directions and 2.5mm slice thickness. On the 3T Trio, TE was 91-93ms, with TR of 7.3-11.6 s and bandwidth of 1395Hz/pixel. On the Skyra, TE was 84ms, with TR of 9.4-9.6 s and bandwidth of 1565Hz/pixel. On the Verio, TE was 91-104ms, with TR of 9.4-14 s and bandwidth of 1395Hz/pixel.

Two methods for OR tractography were assessed (Figure 1). In the hand-drawn method, regions of interest posterior to Meyer’s loop and anterior to tract branching near the calcarine cortex were used. Deterministic streamline fiber tracking with fiber assignment by continuous fiber tracking algorithm was used with minimum FA value of 0.15 and maximum turning angle of 70º. In the automated process, tractography was performed by registering b=0 images to the MNI 152 T1 2mm standard space structural template. Probabilistic streamline fiber tracking was used to isolate white matter tracks between masks of the lateral geniculate nucleus of the thalamus and the occipital pole with exclusion regions in the mid-sagittal image and anterior to the mid-thalamus.

Between 2009-2012, 26 of 50 children (52%) with NF1 and OPG had visual acuity loss. There was no difference in age (p=0.28), tumor location (p=0.14) or DTI parameters (p=0.14) between those with and without visual acuity loss. Hand-drawn tractography methods required 20-25minutes/subject; automated methods were performed with <1minute/subject of operator time. On univariate analysis of all 50 subjects, FA and RD of the optic radiations were significantly different between the two methods. Comparing subjects with and without visual acuity loss, significant differences in RD and MD were found by either method, and in FA using hand-drawn tractography. FA using automated tractography exhibited a trend toward significance (p=0.054) (Figure 2). On multivariable analysis, both methods demonstrated that FA was associated with visual acuity with similar adjusted coefficients and coefficient of determination (R2) (Figure 3).Automated tractography of the optic radiations is a valid method of assessing the association between DTI measures and visual acuity compared to tractography using hand-drawn regions of interest. Because the automated process is constrained to tracts that connect the lateral geniculate nucleus and the occipital cortex, these tracts will differ from previous methods but may better conform to anatomic expectations. Automated tractography of the optic radiations offers a fast, reliable and consistent method of tract identification that is not reliant on operator time and expertise. This method of tract identification may be useful as DTI is included in the routine clinical exam of children with OPG and is developed as a potential biomarker for visual acuity.