The Fuzzy MAD Glioma Conjecture, using Fuzzy C Means to classify Multimodal Apparent Diffusion for glioma stratification

Motivation: Gliomas are a very heterogeneous primary brain tumor. Their heterogeneity lies not only at the microscopic (genetic / biochemical) level, but at the mesoscopic (cellular morphology) level and macroscopic (tissue pathology) level. Goal(s): To better understand the underpinnings of the glioma heterogeneity. Approach: We applied the Multimodal Apparent Diffusion (MAD) method and Fuzzy C Means to multi b‑value diffusion weighted magnetic resonance imaging, up to b‑value of 10K s/mm2, on 54 glioma patients. Results: We discerned 15 normal appearing tissue types and 19 lesion types, including 3 voracious solid tumor, 5 recruiting solid tumor, 3 edema, and 7 b0 attenuated types. Impact: Each stage in the glioma’s progression manifests in unique changes in MAD parameter signatures. The ability to to more precisely understand the heterogeneous microstructure involved can be used to improved diagnosis and prognosis of gliomas.