Effect of inversion time on delayed-enhancement magnetic resonance imaging with and without phase-sensitive reconstruction

Abstract

Purpose: To evaluate the consistency and inversion time (TI) independence of phase-sensitive reconstruction (PSIR) delayed-enhancement (DE) MRI in a clinical setting. Materials and Methods: Mid-ventricular short-axis DE images were acquired in 25 patients using three TIs: 1) optimized to null viable myocardium, 2) 50 msec less than optimal TI, and 3) 50 msec greater than optimal TI. At each TI, images were acquired with PSIR and without magIR. In each image, percent scar was computed as the ratio of nonviable to total pixels in the left ventricle (LV). Results: In the magIR images, percent scar was 23% ± 15% (optimal), 11% ± 11% (short), and 22% ± 15% (long). In PSIR images, percent scar was 25% ± 15% (optimal), 22% ± 15% (short), and 22% ± 14% (long). Percent scar was significantly underestimated in magIR images with short TI, but no statistically significant difference in percent scar was observed at the optimal or long TIs. Conclusion: DE-MRI is a robust imaging technique for clinical use. PSIR provided consistent image quality independently of TI, at least over the range of TIs evaluated in this study. However, neither image quality nor scar appearance in the PSIR images was significantly different from that in the magIR images when TI was at or above the null point of viable myocardium. © 2005 Wiley-Liss, Inc.