Analyzing the Diffusion Patterns for Follow-Up Study of Glioblastoma Multiforme Using Diffusion Tensor Imaging

Hai Li,Zhong Xue,Jiong Xing,Lei Guo,Stephen T. C. Wong
DOI: https://doi.org/10.1109/lissa.2009.4906717
2009-01-01
Abstract:Studying the growth/recurrence of Glioblastoma multiforme (GBM) is very important not only for diagnosis but also for understanding and detecting the recurrence of GBM after surgery. In this paper, a novel DTI-based method is proposed to analyze the recurrence pattern of GBM based on serial Magnetic Resonance Imaging (MRI). After detecting the tumor shapes from T1-weighted images, the diffusion pattern around the tumor can be calculated from the Diffusion Tensor Imaging (DTI) data. This diffusion pattern is then compared with the tumor shapes detected in the follow-up studies, and a quantitative analysis is performed to find the relationship between the morphological changes of the tumor and the diffusion pattern calculated from DTI images. Using the postsurgical longitudinal GBM data acquired from The Methodist Hospital, it has been found that the recurrence patterns of GBM are correlated with the diffusion patterns calculated from DTI images. This finding suggests that the combination of the quantitative measures of both longitudinal morphological and diffusion pattern changes provides more accurate measures about the growth or recurrence of GBM. The proposed method can be used in the follow-up study of GBM as well as in clinical trials of various treatment methods.
What problem does this paper attempt to address?