gms | German Medical Science

Objective: Degeneration of white matter due to tumor growth and edema is a challenge for white matter tractography. Conventional voxel-based methods demonstrate decreased fractional anisotropy in tumorous tissue. Fiber bundle delineation within these areas is considered unreliable. Voxel-based analysis provides an average value of multiple fiber populations within each voxel and does not account for individual fiber integrity differences or crossing fibers. Novel methodology allows now to extract parameters related to fiber density for individual fiber populations within each voxel (so-called Fixels). Our study addresses the hypothesis that in combination with macroscopic morphological information, the new microscopic Fixel-based analysis can provide detailed measures about white matter morphology also within degenerated tumorous brain tissue.

Methods: Fixel-based analysis of within-voxel microscopic fiber density and fiber cross-section was performed. To account for macroscopic white matter morphology, we used Fixel-based morphometry to combine within-voxel microscopic information and macroscopic morphology. For this pilot-study we included four patients with WHO grade III gliomas in the left frontal hemisphere and invasion of the language network and a control group of four subjects without cranial white matter pathologies.

Results: Significant reduction in fiber density and fiber cross-section (p<0.05) was observed in peritumoral areas. Attributing p-values to each Fixel with multiple fiber populations allowed for fiber bundle specific conclusions, which were visualized with group template-based streamlines. Based on this microscopic analysis, conclusive tractography results were obtained in edematous and tumorous regions.

Conclusion: Fixel-based fiber specific white matter analysis allows for estimations of multiple fiber bundles in each voxel and enables detailed appreciation of fiber bundle integrity and morphology, even in tumorous brain tissue.