Glioma Tissue Modeling by Combing the Information of MRI and in vivo Multivoxel MRS
This paper presents a glioma modelization methodand a regression-like model to create a gradually glioma image(GlioIm). Multimodal signal, images of magnetic resonanceimaging (MRI) and in vivo multivoxel MR spectroscopy (MRS)are combined by the regression-like model with spatial resolutionregistration. This modeling method consists of feature models ofglioma such as the signal intensity of MR image and themetabolite changes of MRS, the correlation model noted asmetabolites ratio (MetaR) and the combined regression-likemodel. The estimated GlioIm includes both brain structure andglioma grade information. A nonlinear model is proposed andvalidated in this paper. The testing data is acquired by SiemensTrioTim (3T) and Syngo MR B15 at Beijing Tiantan hospital ofChina. The MRS of three glioma patients, two affected byastrocytoma and one by glioma, and the chemical shift imaging(CSI) reference T2 images were considered in our validationexperiment. The resulting GlioIms are compared with groundtruth provided by neuroradiologists of Tiantan and verified withtheir pathology report. They report that our method and modelare very efficient.
Weibei DOU Aoyan DONG Ping CHI Shaowu LI
Tsinghua National Laboratory for Information Science and Technology Dept.of Electronic Engineering, Neuroimaging Center of Tiantan Hospital Capital Medical University, Beijing, P.R.China Jean-Marc CON
国际会议
成都
英文
1-4
2010-06-18（万方平台首次上网日期，不代表论文的发表时间）