Meeting Banner
Abstract #2934

A Dynamic-Phase Extension for Model-Based Reconstruction of Breast Tumor Dynamic Contrast Enhanced MRI

Benjamin K. Felsted1, Ross T. Whitaker1, Matthias C. Schabel2, Edward V.R. DiBella2

1School of Computing, University of Utah, Salt Lake City, UT, United States; 2Department of Radiology, University of Utah, Salt Lake City, UT, United States


Introduction: We extend the model-based reconstruction method with a physically based linear-phase model that can account for gadolinium field distortions. Methods: Both constant- and linear-phase models were used in reconstructing two 4D breast DCE k-space acquisitions, retrospectively undersampled at R-factors of 1, 4, and 8. Results: Image reconstruction errors correlate spatially with dynamic image phase estimation errors. The errors of the constant-phase model grow fastest as R increases. Conclusion: The new extension can reduce most of the error from phase. The reconstructions have full spatial resolution without the blurring, ghosting, and ringing spatial artifacts typically associated with aggressive undersampling.