Multiparametric Volumetric MR Characterization of Intracranial Lesions

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): MRI acquisitions that map parameters reflecting tissue metabolism and physiology offer considerable potential for improving the yield of diagnostic imaging studies of intra-cranial brain lesions, including better characterization of the type of lesion and increased sensitivity to detect subtle changes of tissue function due to tumor infiltration or response to therapy. A method of particular interest for characterization of brain cancers is MR spectroscopic imaging; however, the techniques currently provided on commercial instruments do not take advantage of recent development in technology and data processing and the potential for new implementations of this imaging modality has yet to be fully evaluated. Additional effort is required to determine the optimal selection of structural and parametric imaging methods for diagnostic studies of brain lesions. This study will develop and make available to other users an efficient implementation of a multiparametric MRI protocol that includes a volumetric "whole-brain" and high spatial resolution MR spectroscopic imaging method that will be integrated onto 3T MRI instruments from Siemens Medical Systems. The MRI protocol will include a novel volumetric and calibrated arterial spin labeling acquisition and diffusion weighted imaging. This combination of advanced imaging modalities will provide optimal sensitivity and spatial coverage for diagnostic MRI studies of intra-cranial masses. The diagnostic efficacy of this multi-parametric and quantitative MRI protocol will then be evaluated for studies of a wide range of brain pathologies in an outpatient setting. Studies will investigate the relative value of structural imaging and parametric imaging protocols, including evaluations of image quality, diagnostic accuracy, and inter- and intra-rater reliability. The data acquired fo this aim will then be used to develop and evaluate computer-aided diagnostic methods using voxel-based tissue classification of the multiparametric, volumetric, and quantitative image data.