(Sunday, 7/14/2019) 1:00 PM - 2:00 PM
Room: 221CD
Purpose: It is an emerging approach to combine multiple contrasts into a single analysis. This approach would allow an exploration of the T2-relaxation-diffusion correlation in brain tumors to better understand tumor physiology than using diffusion or T2-relaxation MRI alone. Gross tumor volume on T2 FLAIR images (GTV_FLAIR) usually contains tissue mixture and is difficult to be analyzed by a single technique. In a previous study, we have explored the T2-relaxation-diffusion correlation of glioblastoma (GBM) and normal tissue, and shown significant differences between 2 diffusion coefficients and T2 values among 5 tissue types. This research aims to combine a cluster classification method with the T2-relaxation-diffusion correlation to demonstrate T2 variation in GBM.
Methods: Twelve patients with histologically diagnosed GBM had DW images with 11 b-values from 0 to 2500 s/mm² at 2 TEs (93 ms and 113 ms) on a 3T scanner. GTV_FLAIR of each patient was pre-defined by experienced physicians and classified to four classes using diffusion weighted (DW) image intensities with fuzzy c-means. Here, we considered fast (D(f)) and slow (D(s)) diffusion coefficients and corresponding T2 values of T2(f) and T2(s).
Results: Two classes had both T2 values and diffusion coefficients close to those from hypercellular tumor volumes (TV(HCV)) that were determined previously. Third class had the T2 value associated with the fast diffusion coefficient distant from one from TV(HCV) and close to normal tissue. The fourth class contains most CSF and fluid from the surgical cavity.
Conclusion: Large T2 variations associated with fast diffusion coefficients in GBM are revealed by the analysis of the T2-relaxation-diffusion correlation. This might not be able to be achieved by analyzing T2 relaxation and diffusion MRI alone. The T2-relaxation-diffusion correlation measurement has the potential to identify GBM in FLAIR abnormality.
Funding Support, Disclosures, and Conflict of Interest: This work is supported in part by a grant of NIH/NCI 1U01CA183848
Keywords
Diffusion, MRI, Image Analysis
Taxonomy
Contact Email
