Rapid high‐resolution T₁ mapping by variable flip angles: Accurate and precise measurements in the presence of radiofrequency field inhomogeneity

Abstract

Rapid 3D mapping of T 1 relaxation times is valuable in diverse clinical applications. Recently, the variable flip angle (VFA) spoiled gradient recalled echo approach was shown to be a practical alternative to conventional methods, providing better precision and speed. However, the method is known to be sensitive to transmit field (B 1 ؉ ) inhomogeneity and can result in significant systematic errors in T 1 estimates, especially at high field strengths. The main challenge is to improve the accuracy of the VFA approach without sacrificing speed. In this article, the VFA method was optimized for both accuracy and precision by considering the influence of imperfect transmit fields, noise bias, and selection of flip angles. An analytic solution was developed for systematic B 1 ؉ -induced T 1 errors and allows simple correction of T 1 measurements acquired with any imaging parameters. A noise threshold was also identified and provided a guideline for avoiding T 1 biases. Finally, it was shown that three flip angles were the most efficient for maintaining accuracy and high precision over large ranges of Rapid and accurate measurement of the longitudinal, or T 1 , relaxation time has long challenged MRI scientists but remains an important goal because of its clinical relevance across a diverse range of applications. Many of these applications, however, including dynamic contrast-enhanced studies of cancer, perfusion studies of muscle, diagnosis of neurologic disorders such as multiple sclerosis and epilepsy, and enhanced tissue discrimination for image-guided procedures, require low-noise, high-resolution mapping over a large volume. These requirements cannot be achieved in a clinically acceptable time frame (Ͻ30 min) using conventional methods of inversion-or saturation-recovery (1-3). An alternative method (4) involves determining T 1 from a set of two or more spoiled gradient recalled-echo (SPGR) images acquired with varied flip angles. Advantages of this approach include low power deposition compared to spin echo techniques and low spatial distortion compared to rapid echo planar imaging. Most importantly, the accuracy has been shown to be similar to that achieved with conventional and accelerated techniques but with a significant reduction in imaging time (5,6).A dominant source of error in the variable flip angle (VFA) approach is inaccurate knowledge of the flip angles due to transmit field B 1 ϩ inhomogeneity. Although the impact on T 1 accuracy is known (7,8), correction of measured T 1 has been reported in only a few studies (9,10), mainly due to the complexity and long scan time requirements of mapping the B 1 ϩ field. A second source of systematic error is noise-induced bias. The effect on T 1 measurements is generally subtle and may be the reason it has not, to our knowledge, been previously reported. However, these errors can be appreciable below a certain signal-tonoise (SNR) threshold, particularly with multiple angles, and one must identify and image at a suitable SNR level to ensure accurate measureme...

Benign prostatic hyperplasia: Evaluation of T1, T2, and microvascular characteristics with T1‐weighted dynamic contrast‐enhanced MRI

“…Improvements in the imaging protocol from a previous study of prostate cancer (11,14) were assessed in terms of the increase in precision calculated using a bootstrap estimate of uncertainty. The variable flip angle method for measuring T 1 relies on accurate knowledge of the applied flip angles, which can be quite different from the angle intended due to B 1 inhomogeneity (28). The IRTFE sequence has previously been shown to give an accurate measurement of T 1 while being largely insensitive to B 1 inhomogeneity (29,30) This method has also been shown here to be highly reproducible (Յ4%).…”

Quantification of renal perfusion: Comparison of arterial spin labeling and dynamic contrast‐enhanced MRI

“…Kidney tissue uptake curves were then generated. Preinjection T 1 values were calculated from the variable flip angle data and corrected for B 1 variations (24). Mean renal cortex and mean medulla tissue curves were extracted, as well as pixel-wise cortical tissue curves.…”

Optimized T1- and T2-weighted volumetric brain imaging as a diagnostic tool in very preterm neonates

“…The inclusion criteria for these cases included germinal matrix hemorrhage (GMH) grade I, small ischemic focus and no abnormalities at all. T1 parametric maps were produced using a linear least squares solution according to the methodology presented by Cheng and Wright [39]. For accurate flip angle calculation, a rapid B 1 + mapping sequence was employed.…”