Meeting Banner
Abstract #0826

Histogram Analysis of Fractional Anisotropy & Mean Diffusivity: Distinguishing Tumefactive Demyelinating Lesions from High-Grade Glioma Using DTI

B. LIU1, X. LIU2, W. TIAN3, T. Zhu4, S. Ekholm, J. ZHONG4

1Electrical and Computer Engineering , University of Rochester, Rochester, NY, United States; 2Department of Radiology, University of Rochester, United States; 3Department of Radiology, University of Rochester, Rochester, United States; 4Imaging Sciences, University of Rochester, Rochester, NY, United States

It is of significant clinical importance to differentiate between Tumefactive demyelinating lesions (TDLs)and high-grade gliomas(HGGs) to avoid surgical biopsy. However, this goal is difficult to achieve by conventional MR imaging alone. The objective of our study is to determine retrospectively whether quantitative histogram analysis of FA and MD values can be helpful in distinguishing between TDLs and HGGs. We compare the FA value and MD value for both enhanced Tumefactive demyelinating lesions(TDLs) and enhanced high grade glioma(HGGs) with the matched location. Results showed that TDLs and HGGs could be differentiated by using the histogram method.

Abstract PDF Poster PDF

Keywords

able achieve acquisition acute alone analyses anisotropy appearance assess avoid axial biopsy brain calculation clinical collected complete computer confirmed contrast corrected defined determine diagnosis differentiate differentiated difficult diffusion diffusivity disseminated distinguish distinguishing distortion distribution distributions divided eddy electrical engineering enhanced enhancement entity evaluate features focal fractional goal grade height heights helpful histogram histograms importance included indicated intermediate investigated leads lesion lesions like location make many maps mass matched matrix measured measuring medical mimicking morbidity nonspecific normalized objective occur offered often opportunity oxford package pathological pathology patient patients peak percentile plot plots portions post processing produce produced proved quantify quantitative radiology respectively retrospective review risk scanner scanning sciences sclerosis separate shared significantly slice solitary statistical steps studied studies subject subjects suggest surgical systems table tensor tumor tumors type types typical usually whether