TrueFISP: assessment of accuracy for measurement of left ventricular mass in an animal model

Abstract

Purpose: To test the accuracy of a high performance true fast imaging with steady-state precession (TrueFISP) pulse sequence for the assessment of left ventricular (LV) mass in a large animal model on 1.5-T scanners.

Materials and methods: We imaged dogs (N = 10) on a clinical 1.5-T clinical scanner using electrocardiogram (ECG)-gated TrueFISP. In all animals, contiguous segmented k-space cine images were acquired from base to apex (in-plane resolution 1 x 1 mm(2), slice thickness 5 mm, TR = 4.8 msec, TE = 1.6 msec) during repeated breath-holds. In nine of the 10 animals, single-shot images gated to end-diastole were also acquired from base to apex in a single breath-hold (in-plane resolution 1 x 1 mm(2), slice thickness 5 mm, TR = 3.2 msec, TE = 1.6 msec). After imaging, animals were killed, the left ventricle was isolated, and the true mass of the left ventricle (free wall and septum) was determined. Independently, two observers blinded to the post-mortem results computed LV masses based on analysis of the magnetic resonance (MR) images.

Results: Comparison of the computed LV mass using TrueFISP to the actual mass showed excellent agreement. Cine-systole was the most accurate technique (mass = 98.6% +/- 4.5% actual, bias = 1.2 +/- 3.4 g) followed by cine-diastole (mass = 97.9% +/- 5.3% actual, bias = 1.8 +/- 4.1 g) and single shot (mass = 94.7% +/- 7.9% actual, bias = 4.2 +/- 6.3 g). Inter- and intra-observer variabilities were low (5.8% +/- 7.1% and 0.4% +/- 4.8%, respectively).

Conclusion: We conclude that TrueFISP imaging is an accurate, rapid method to determine ventricular mass. In single-shot mode, TrueFISP requires only one breath-hold to estimate the mass of the heart within 6% of the actual value, whereas the segmented k-space implementation measured LV mass to within 3% of the true value.