Meeting Banner
Abstract #2627

Two Non-Linear Parametric Models of Enhancement for Breast DCE-MRI That Can Be Fitted using Linear Least Squares

Andrew Mehnert1, Michael Wildermoth1, Stuart Crozier1, Ewert Bengtsson2, Dominic Kennedy3

1School of ITEE, the University of Queensland, Brisbane, Qld, Australia; 2Centre for Image Analysis, Uppsala University, Sweden; 3Queensland X-Ray, Greenslopes Private Hospital, Greenslopes, Australia


Two non-linear empirical parametric models are proffered for use in quantitatively characterizing contrast enhancement in dynamic contrast enhanced MRI of the breast: linear-slope and Ricker. The advantage of these models over existing pharmacokinetic and empirical models is that they can be fitted using linear least squares which means that fitting is quick, there is no need to specify initial parameter estimates, and there are no convergence issues. An empirical evaluation of the goodness-of-fit of these two models relative to the Hayton and the simplified gamma-variate model is also presented.