Cerebral Gliomas: Diffusional Kurtosis Imaging Analysis of Microstructural Differences

Abstract

To characterize the non-Gaussian diffusion patterns of cerebral glioma microstructure with respect to the different glioma grades by using a new method called diffusional kurtosis (DK) imaging. In this study with institutional review board approval and patient consent, diffusional measures of mean kurtosis (MK), fractional anisotropy (FA), and apparent diffusion coefficient (ADC) were compared prospectively. Data were normalized to the contralateral white matter. A Mann-Whitney test was used to compare each histologic glioma subtype regarding the diffusion measurements. Receiver operating characteristic curves were used to test for the parameter with the best sensitivity and specificity for glioma grade discrimination. In 34 patients with cerebral gliomas (five World Health Organization [WHO] grade II astrocytomas, 13 WHO grade III astrocytomas, and 16 WHO grade IV glioblastomas multiforme), significantly different diffusion patterns were found among the three glioma groups. MK values increased with higher glioma malignancy, whereas ADCs tended to decrease with higher malignancy; FA values did not differ significantly among tumor groups. Significant differences between astrocytoma grades WHO II and WHO III were demonstrated only by DK values. Area under the receiver operating characteristic curve was highest for normalized MK (0.972) during testing to discriminate between low- and high-grade gliomas. This study demonstrates specific diffusion patterns for low- and high-grade gliomas, showing that DK imaging is able to depict microstructural changes within glioma tissue and is able to help differentiate among glioma grades. (c) RSNA, 2010.