Neville Gai PhD, Presenter: Nothing to Disclose

Marcelo Souto Nacif MD, PhD, Abstract Co-Author: Nothing to Disclose

Evrim Bengi Turkbey MD, Abstract Co-Author: Nothing to Disclose

Songtao Liu MD, Abstract Co-Author: Nothing to Disclose

Christopher Sibley, Abstract Co-Author: Nothing to Disclose

David A. Bluemke MD, PhD, Abstract Co-Author: Nothing to Disclose

Myocardial fibrosis can be detected by cardiac MRI, but interstitial fibrosis parameters have been shown to vary widely between patients. The purpose of this study was to determine if additional information available can allow improved interpatient evaluations.

Previously an analytical model, which corrects for T1 variation resulting from differences in dose, post-contrast time and glomerular filtration rate (GFR), was presented. The model determines the correction using mean extracellular volume fraction (Ve) and blood plasma-extracellular space exchange rate (K2) by curve fitting to post-contrast T1 values determined at several time points. Using blood T1 values along with the hematocrit (Hct) can allow for independent determination of Ve from Ve = (1-Hct)*ρ*(ΔR1myo/ΔR1blood)-Vp (ρ=1.05, Vp=0.045). The number of unknowns in then reduced resulting in better determination of K2. In addition, correlating K2 to post-contrast myocardial T1 values allows for a better estimate of K2 to be used for each patient resulting in improved accuracy of correction. The changes described above were tested for three different contrast agents (Magnevist®, Multihance® and Ablavar®) for T1 values obtained at several time points in 21 volunteers. The new changes were tested for goodness of fit and for predicting T1 values at a time point from values available at different time points.

Goodness of fit to measured T1 values was significantly improved (p<0.01) with the new approach when compared with the previous method. Overall Pearson correlation of K2 to (Dose(D)×T1myo) was significantly strong (p<0.02). Linear fits for (D×T1myo) to K2 determined for each of the three contrast agents was used in the correction. Mean error between the predicted T1 value and the observed T1 value (available from kinetic studies) at a given time point was found to be 1.86% and 2.8% with and without the changes, respectively.  

A model where Ve and K2 (both physiological parameters that vary between subjects) can be tailored for each T1 measurement improves accuracy of correction. Accurate correction allows for better comparison between post-contrast T1 values in turn improving reliability of any  inferences for diffuse fibrosis made based on T1 values.

Comparison of T1 times between different patients requires accurate correction for patient specific factors such as contrast dose, GFR and different post-contrast times.

Gai, N, Nacif, M, Turkbey, E, Liu, S, Sibley, C, Bluemke, D, Estimate of Myocardial T1 to Assess Interstitial Fibrosis: Use of Kinetic Parameter Relationships and Blood Pool Data for Improved Correction of Interpatient T1 Variation.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12026713.html