Meeting Banner
Abstract #0616

An Image Similarity-Guided Correspondence Correction for Voxel-Wise Analysis Applied to MR Imaging of Glioblastoma Multiforme Acquired Pre- And Post-Chemoradiotherapy

Jeremy David Hoisak1,2, Eng-Siew Koh1,3, Eugene Yu4, Andrea Kassner4, Normand J. Laperriere1,3, Cynthia Mnard1,3, David A. Jaffray1,2

1Radiation Medicine Program, Princess Margaret Hospital, Toronto, Ontario, Canada; 2Medical Biophysics, University of Toronto, Toronto, Ontario, Canada; 3Radiation Oncology, University of Toronto, Toronto, Ontario, Canada; 4Medical Imaging, University of Toronto, Toronto, Ontario, Canada


Response assessment with a voxel-wise analysis of serial image change has advantages over conventional tumor measurements, but is susceptible to uncertainties from inconsistent voxel correspondences between scans arising from a dynamic tumor morphology. A correspondence correction method based on a metric of voxel similarity was applied to a functional diffusion map (fDM) analysis of co-registered T1-weighted and diffusion-weighted images of glioblastoma multiforme acquired pre- and post-chemoradiotherapy. The correction resulted in a statistically significant alteration in the quantification of apparent diffusion coefficient (ADC) change pre- and post-therapy, and has the potential to improve the accuracy of subsequent determinations of therapy outcome.