Comparison of T1 estimation techniques in cardiac MRI

Detection and prediction of acute heart transplant rejection with the myocardial T2 determination provided by a black-blood magnetic resonance imaging sequence

OBJECTIVES

This study aimed to determine whether the myocardial T2 relaxation time, determined using a black-blood magnetic resonance imaging (MRI) sequence, could predict acute heart transplant rejection.

BACKGROUND

The use of black-blood MRI sequences allows suppression of the confusing influence of blood signal when myocardial T2 is calculated to detect myocardial edema.

METHODS

A total of 123 investigations, including cardiac MRI and myocardial biopsy, were performed 8 +/- 11 months after heart transplantation. Myocardial T2 was determined using an original inversion-recovery/spin-echo sequence.

RESULTS

A higher than normal T2 (> or = 56 ms) allowed an accurate detection of the moderate acute rejections evidenced at baseline biopsy (> or = International Society for Heart and Lung Transplantation grade 2): sensitivity, 89% and specificity, 70% (p < 0.0001). T2 was increased in grade 2 (n = 11) compared with grade 0 (n = 49, p < 0.05), grade 1A (n = 34, p < 0.05) and grade 1B (n = 21, p < 0.05); T2 was further increased in grade 3 (n = 8) compared with grade 2 (p < 0.05). In addition, in patients without rejection equal to or greater than grade 2 at baseline, a T2 higher than normal (> or = 56 ms) was correlated with the subsequent occurrence of equal or greater than grade 2 rejection within the next three months: sensitivity 63% (12/19) and specificity 78% (64/82) (p = 0.001).

CONCLUSIONS

Myocardial T2 determined using a black-blood MRI sequence, is sufficiently sensitive to identify most of the moderate acute rejections documented with biopsy at the same time, but is also a predictor of the subsequent occurrence of such biopsy-defined rejections.

T2 maximum likelihood estimation from multiple spin-echo magnitude images

An optimal maximum likelihood (ML) method is described for an unbiased estimation of monoexponential T2 from magnitude spin-echo images. The algorithm is based on a Gaussian assumption of noise distribution. The validity of this assumption was checked by a statistical chi 2 test on spin-echo and fast low-angle shot surface coil images. Monte-Carlo simulations of magnitude data showed that the ML estimate standard deviation is lower than that produced by a weighted least-squares fitting on signal logarithm. Correction schemes are proposed to reduce bias deriving from magnitude reconstruction. The variance of the ML estimate converged rapidly toward the theoretical algebraic expression of the Cramér-Rao lower bound.