Meeting Banner
Abstract #0548

Assessment of Abnormal ADC Matched Voxels with DCE Parameters for Characterization of Prostate Cancer at 3T

Fiona M. Fennessy1, 2, Andriy Fedorov1, Sandeep N. Gupta3, William M. Wells1, Robert V. Mulkern1, 4, Clare M. Tempany1

1Radiology, Brigham and Women's Hospital, Boston, MA, United States; 2Radiology, Dana Farber Cancer Institute, Boston, MA, United States; 3GE Global Research Center, Niskayuna, NY, United States; 4Radiology, Children's Hospital Boston, Boston, MA, United States

We sought to determine if Ktrans and ve differed within ADC-matched prostate tumor voxels, confirmed by pathology at radical prostatectomy. A deformable registration technique was used to recover image distortions on DWI, registered to the DCE series. We found a significant difference between tumor and normal tissue for ADC, Ktrans , and ve. Using Logistic regression analysis, Ktrans has a contributory role to ADC for tumor detection. However, weak correlations between both Ktrans and ve and the ADC values, underscoring the different pathophysiological processes assessed with DWI and DCE, and need for a multiple-parameter strategy approach for prostate tumor assessment.

Abstract PDF Presentation Video

Keywords

abnormal adherence affected aligned allow alone arterial artifacts assessed assessment association assumed available basis benign black cancer children circles clearly cohort combined concentration confirmed conjunction considered constant contoured contrast contribute correlate correlation correlations custom days deformable demonstrating derived detailed detection determine differed differentiate differentiating difficult diffusion done extracellular fact failure fast fraction function global gradient green heterogeneity highlight hypo identification improve injection intensity inter investigate investigated like location locations logistic malignancy maps matched matrix meters mine minutes model noon norm observation offers often outlined outlining parametric particularly pass pathology patients peak planar play population post processes prostate protocol quantitative radical radiology rank reader registered registration regression removed repeat response review role scanner score series signed slab slicer slices sought space spacing spatial spoiled stud subset suffers sufficient suspicious table tissue tool trace trans tumor turn underwent volume ward weak ween wise women years zero zone