Improved T1 quantification using post-Gd contrast variable flip angle data

Introduction: In dynamic contrast-enhanced (DCE-) MRI studies, it is necessary to obtain estimates of initial (i.e. native) and dynamic T1 values in order to quantify contrast agent concentrations, [CA]. The variable flip angle (VFA) method [1], which employs spoiled gradient echo data acquired using at least two different flip angles, is a T1 and M0 (a variable including the proton density, gain factor and echo term) estimation method that is commonly employed in DCEMRI studies. The optimal choice of flip angles depends on the repetition time and the range of T1 values present in the imaging sample [2,3]. Flip angle optimisation in DCE-MRI studies is a challenge because the observed T1 range is large, especially in blood, which has a large peak [CA]. Furthermore, the choice of flip angle for the dynamic acquisition needs to ensure that signal linearity is maintained over a range of [CA]s for different initial T1s. Flip angle optimisations have shown that the larger the maximum T1 in the range, the smaller the optimal flip angles are [2,3]. For greater signal linearity with [CA], however, a larger dynamic flip angle is desirable. When designing a DCE-MRI protocol, the choice of flip angles needs to reflect a trade-off between the two requirements. At the end of the dynamic study, maximum T1 in the range is shortened (relative to pre-contrast) as a result of CA uptake and the signal-to-noise ratio is improved [4]. In theory, therefore, it would be possible to obtain relatively precise T1 estimates using the VFA data obtained at the end of the DCE-MRI study (i.e final T1), and also M0 estimates. The estimate can be propagated backward through the dynamic data to generate dynamic T1 curves, thus, enabling an estimation of the initial T1. In this study, we compared initial T1 values estimated using the described, ‘post-Gd converter’, method with those estimated using the normal method.