Meeting Banner
Abstract #0841

DSC-MRI Perfusion in High Grade Cerebral Tumors: Solution of Leakage Effects in a Simple Clinical Image Processing Protocol.

Guy Umberto Poloni1, Enrico Marchioni2, Giovanni Magenes3, Gloria Castellazzi4

1Neuroradiology, Fondazione "Istituto Neurologico Casimiro Mondino" IRCCS, Pavia, PV, Italy; 2Neurology, Fondazione "Istituto Neurologico Casimiro Mondino" IRCCS, Pavia, PV, Italy; 3Bioengineering Department, Universita' degli Studi di Pavia, Pavia, 27100, Italy; 4Bioengineering Department, Universita' degli Studi di Pavia, Pavia, PV, Italy

DSC-MRI represents the widespread approach in neuroradiology for the PWI of cerebral tumors, limited in its efficacy by the breakdown of the BBB associated to high grade cerebral tumors. A multi-parametric approach was adopted to overcome technical limitations of the technique on 30 consecutive untreated patients affected by glioma of grades ranging from II to IV. A voxel-wise analysis of the perfusion T2* relaxivity curves, adopting a higher number of parameters than standard and an increasing level of sophistication of the pre-analysis, was performed demonstrating an improved sensitivity to cerebral hemodynamic alterations, particularly in regions where BBB injury occurs.

Abstract PDF

Keywords

absence accuracy accurate affected agreement among appropriate assess assumption axially barrier behavior better bioengineering blood bolus brain breakdown cerebral characterizes classification clinical completely concentration consent considered consistently contrast correlate correlated correlation curves dataset decomposition defined dependence diagnosis differentiate difficult discriminate displayed disruption dynamic evaluate evaluated evaluation exam feasibility females filtering findings flair flow former grade grades grading gradually heterogeneity identify improve included increasing independently indicators injection injury intensity interval kinetic leakage lesion literature local males maps materials maximize methodological model negative observations occupied occurs often oversampling parametric particular particularly passage pathology patients peak perfusion position post preliminary prepared presence presto principles prior processing protocol quantitative radiology ranging realignment reduces relation reliability reported requires respectively sample scheme schemes score sensitivity series shape signed significance significantly similarly sophisticated sophistication specificity start step steps stress susceptibility tissue tissues together tracking train transit treated tumor tumors untreated utility volume volumes wash washout widespread width wise