Meeting Banner
Abstract #0971

Repeatability of the Perfusion MRI Brain Tumor Vasculature Sensitive Biomarker, Arterio-Venous Overlap (AVOL) in Recurrent Brain Tumor Patients with Two Baseline Imaging Scans

Peter S. LaViolette1, Mitchell Daun2, Melissa A. Prah1, Kourosh Jafari-Khouzani3, Pavlina Polaskova3, Elizabeth R. Gerstner4, Steven M. Stufflebeam3, Kathleen M. Schmainda1

1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States; 2Neurology, Medical College of Wisconsin, Milwaukee, WI, United States; 3Radiology, Massachusetts General Hospital, Charlestown, MA, United States; 4Neurology, Massachusetts General Hospital, Boston, MA, United States

A new biomarker of brain tumor vasculature derived from independent component analysis (ICA) applied to DSC MRI data has been recently described. It has been shown to occur preferentially in tumor, and is predictive of response to bevacizumab. This study varies the number of ICA components modeled to determine what number is most appropriate for modeling AVOL. The study was done in 27 patients with two scans days apart prior to therapy. We also compared simultaneously acquired SE and GE data. We find that modeling 4 or 5 components results in highly repeatable GE maps, while SE maps of AVOL are less repeatable.

Abstract PDF

Keywords

acquisition acquisitions additional addresses administered advancing apart appears arterial audience best brain brought characteristics clinicians collected college combined combining component components consisted contrast correction created creation dataset days demonstration derived deriving describes determine developmental differ drawn dual dynamic early enhancement enhancing equation funding general generally generated gradient greater healthier highly hospital ideal identical implemented independent index indicates initial leakage library manually maps masked maximally maximized measures medical melodic minimal minimize modeled modeling motion necessary neoplastic neurology novel occurs onset optimizes overall overlap overlapping patient patients perfect perfusion peter place planar population prefigure prescription press prior probabilistic processed processing proper proportions protocol questions radiology recent recruited recurrent registering removing repeatability repeatable repeated report represent represented scanned scientists sensitive separate separately separates session sessions significantly slice smith sorted space spin stages statistically suggesting survival susceptibility system target thresholded took tool trans treatment tumor twice unique venous versus visit visualize visually workshop