Voxelwise Analysis of Imaging Response to Therapy in Neurooncology

• Activity Number: 629
• Type: Contributed
• Date/Time: Thursday, August 5, 2010 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Mengye Guo*+ and Jeffrey Yap and Annick D. Van den Abbeele and Nancy U. Lin and Armin Schwartzman
• Companies: Dana-Farber Cancer Institute and Dana-Farber Cancer Institute and Dana-Farber Cancer Institute and Dana-Farber Cancer Institute and Harvard School of Public Health
• Keywords: EM ; Mixture distribution ; Multivariate ; Multiple testing

Abstract

In neurooncology, positron emission tomography(PET) has been used to evaluate the response of brain tumors to treatment by measuring the change in glucose metabolism expressed as image intensity between a baseline and a follow-up scan. This is often accomplished by manually drawing a region of interest and computing the intensity change within the region over time, which is subject to observer variability. Moreover, one needs to adjust for the background varying over time caused by normal brain metabolism or global drug effects. We propose an analysis pipeline for automatically detecting the change while adjusting for the background. The complete analysis includes image registration, segmentation, a hierarchical model for background adjustment and voxelwise statistical comparisons. Our method has been applied to a neurooncology study and generated promising and interpretable results.