Improvement Grading Brain Glioma Using T2 Relaxation Times and Susceptibility-Weighted Images in Mri Publisher



Tavakoli MB¹ ; Khorasani A¹ ; Jalilian M² Show Affiliations
Source: Informatics in Medicine Unlocked Published:2023

Abstract

Developing a new non-invasive, and accurate method for glioma grading for glioma treatment and management is very important. Magnetic resonance imaging (MRI) has many applications in the field of neuroimaging. The study aims to create a precise way to improve glioma grading using different MRI image weights. There were 47 participants in this study. A 1.5 T MRI scanner acquired T2 multi-echoes (T2 ME) with eight echoes for T2 maps reconstruction, Susceptibility-Weighted Imaging (SWI), and T1 post-contrast. T2 maps were reconstructed from T2 ME data using homemade MATLAB code. To compare T2 relaxation times and intratumoral susceptibility signal intensity (ITSS) between the high-grade gliomas (HGGs) and low-grade gliomas (LGGs) and other analyses, we used SPSS software. There was a significant difference in T2 relaxation times and ITSS between the HGGs and LGGs. The average T2 relaxation times in HGGs and LGGs were 0.12 and 0.16 s, respectively. Further analysis showed that the ITSS in HGGs is significantly greater than in LGGs. There was a significant negative correlation between T2 relaxation times and ITSS in gliomas. It is interesting to note that combining T2 relaxation times, ITSS, and tumor enhancement status data with a multivariate binary logistic regression model can increase the area under the curve (AUC) to 0.94 for glioma-grade classification. We recommended the clinical use of combined T2 maps, SWI, and T1 post-contrast image-weights data with a multivariate binary logistic regression model for improving glioma grading. © 2023 The Authors