Towards Improving Tumor Boundary Identification in Murine Models of Glioma using Cerebral Blood Volume Maps

Kathleen E. Chaffee¹, Jeff R. Anderson¹, Joshua S. Shimony¹, G. Larry Bretthorst¹, Joseph J. H. Ackerman¹, Joel R. Garbow¹

Accurate, non-invasive determination of tumor boundaries remains a significant challenge for neurosurgeons, with important implications for patient management. MR contrast enhancement, which reflects the breakdown of the blood brain barrier, often serves as a surrogate marker for actively growing tumor. Here we utilize in vivo LAIF-DSC-derived CBV maps as a marker for delineating tumor margins in murine models of glioma. The LAIF-DSC method permits the calculation of perfusion-based parametric maps without requiring independent measurement of an arterial input function. Tumor margins, estimated by histogram analysis and thresholding of the CBV parametric maps, correlate well with histology.