White matter characterization in regions of edema surrounding meningioma brain tumor using diffusion MRI: A comparative study of DTI and NODDI.

White matter (WM) tract detection is critical in presurgical planning of tumor resection; however, standard-of-care diffusion tensor imaging (DTI) often fails to characterize white matter tracts through regions of edema. This is because the presence of edema has the effect of increasing the isotropic volume fraction within a voxel and thus marginalizing the anisotropic volume fraction associated with white matter presence and directionality. More recent biophysical models of diffusion, such as neurite orientation dispersion and density imaging (NODDI), account for isotropic and anisotropic volume fractions within voxels by compartmentalizing the diffusion signal based on an assumed tissue microenvironment, e.g., "free water" (cerebrospinal fluid (CSF), interstitial fluid (ISF), edema), "intra-neurite", and "extra-neurite" tissue, as a sphere, stick, and tensor, respectively. We hypothesize that a low fractional anisotropy (FA), low orientation dispersion index (ODI) value and high fractional isotropic volume (FISO) would be observed in white matter regions containing edema but a high FA, low ODI value and low FISO would be observed in healthy-appearing contralateral white matter. In our study, we test this hypothesis using multi-shell diffusion MRI data collected from patients bearing meningioma brains tumors. Brains bearing meningioma tumors are selected in this study as meningiomas rarely invade the brain parenchyma and we can thus assume that our analyses of edematous regions are not confounded by infiltrating tumor cells. Here, we show that NODDI-based characterization of white matter is more sensitive than that of standard-of-care DTI through regions of edema. Future studies will focus on implementation of biophysical model-based tractography in cases of glioma and translation of biophysical model-based tractography to the operating room.