Quantitative in-vivo imaging of tumor microenvironments

Abstract

Tumor hypoxia, which develops heterogeneously in locally advanced tumors is known to affect radiation sensitivity and development to metastases. In vivo knowledge of hypoxia distribution in solid tumors provides prognostic information and can be potentially used for input for dose escalation in radiation therapy. Tumor hypoxia results from a mismatch between supply and consumption of oxygen in a tumor. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is well known to provide permeability/perfusion information of solid tumors and may provide surrogate information regarding tumor hypoxia. In this study, (1) DCE-MRI data with the injection of Gd-DTPA was analyzed with Gaussian mixture model (GMM) based classification to verify regions of perfused, hypoxic, necrotic areas in a prostate rat tumor model. The results of pattern recognition on the DCE-MRI show the feasibility on delineation of tumor microenvironments. (2) To increase the spatial/temporal accuracy of such classification, a compressed sensing algorithm is used to enhance the quality of DCE-MRI uptake curves.

Keywords

• Tumor microenvironments
• hypoxia
• DCE-MRI
• compressed sensing