Meeting Banner
Abstract #0081

Multiparametric Microvascular MRI: A Cluster Approach to Characterize Glioma

Nicolas Coquery1, 2, Clment S. Debacker3, 4, Rgine Farion, 25, Chantal Rmy1, 2, Olivier Francois6, 7, Emmanuel Luc Barbier, 28

1U836, INSERM, Grenoble, France; 2Universit Joseph Fourier, grenoble, France; 3Universit Joseph Fourier, Grenoble, France; 4Bruker Biospin MRI, Ettlingen, Germany; 5Grenoble MRI Facility IRMaGe, Grenoble, France; 6TIMC-IMAG laboratory, UMR5525, CNRS, La Tronche, France; 7Universit Joseph Fourier, La Tronche, France; 8INSERM U836, Grenoble, France

Multiple parameters of tumor microvasculature can be assessed with MRI. The accumulation of physiologically linked information might not be readily interpretable. To address this concern, we propose a cluster-based approach to highlight independent microvasular features within tumor. This strategy could distinguish two glioma models based on their cluster composition. This approach might be useful for solid tumor diagnosis as well as for the localization of pathological areas within the tumor that might become resistant to treatment.

Abstract PDF Presentation Video

Keywords

accumulation agreement analyzed animals apparent appears arterial assessed association become behavior black blood blue calculation cells cerebral characteristic characteristics characterize characterized chelate choice cluster clusters code color composed compositions computed constant continuous correspond corresponds cortex criterion crosses curve days decrease decreases define delay depicted described detect determined deviation diagnosis diffusion distribution drawn duration ease evolution example excluded expectation facility features five flow fraction gathered great green growth healthy heterogeneously highlights impact impacts implemented injection inoculated inoculation integrity interpretable intravenous journal known labeling laboratory largest least links look mainly manually mapped maps materials maximization metabolic might mixture model modeling models observations optimal orange organization oxygen package pairs physiological pixels potential preserves previously prognosis pronounced purple qualitative radiology rats readily reduced reduction related represent resistant saturation several spatial spin standardized statistical structure suggests surface surprising tight tissue treatment tumor tumors types vascular volume wall wealth whereas yellow