Model‐based T₁ mapping with sparsity constraints using single‐shot inversion‐recovery radial FLASH

Abstract

The comprehensive model-based reconstruction with L1 regularization offers rapid and robust T mapping with high accuracy and precision. The method warrants accelerated computing and online implementation for extended clinical trials. Magn Reson Med 79:730-740, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Saturation-Recovery Myocardial T1-Mapping during Systole: Accurate and Robust Quantification in the Presence of Arrhythmia

“…This trade-off led to decreased precision in systolic T 1 -maps of healthy volunteers . Future studies will focus on employing advanced acceleration techniques 19 , 20 , potentially exploiting the interdependence between the baseline images 21 , in order to mitigate this loss in precision. In systole, moreover, a smaller slice thickness of 6 mm (versus 8 mm in diastole) was chosen to avoid partial volume effects with the high signal from the adjacent blood pool, which is particularly important in systolic imaging as the cardiac long axis is substantially shortened during the contraction.…”

Improvement of Fast Model-Based Acceleration of Parameter Look-Locker T1 Mapping

“…In order to get better time of computational we decided to choose a simple and efficient method of model fitting. Other solutions, in some cases are more efficient but are also numerically advanced, and require more time to reconstruct the final image [16][17][18][19][27][28][29]33]. In the literature we did not find any benchmark that enables us to compare methods of T 1 mapping reconstruction and quality of results of different works for similar data.…”

“…The basic IR-MAP algorithm [14] requires approximately 100 s (reported 90 s for fitting procedure and remaining part for reinserting) to evaluate a single iteration, which for an assumed 50 iterations gives approximately 85 min of the whole iteration process and data preparation for a single slice. Acceleration using GPU and implementation of some methods in C/CUDA (Compute Unified Device Architecture) generates a time complexity of 10-20 min [19,33] and in the range from minutes to hours depending on data size [18]. Fast multi-slice method for T 2 mapping [29] calculates 50 slices on an office computer within 7 h, which gives approximately 9 min per slice or alternatively rapid T 1 quantification [28] reports reconstruction time of approximately 10 min per slice.…”

“…The proposed algorithms require many iterations of fitting and computational time is higher than one hour for a single slice on a home computer without central processing unit (CPU) and graphics processing unit (GPU) acceleration. The series of different methods were presented by Wang et al in the form of regularized nonlinear inversion (NLINV), conjugate gradient (CG) along with pixel-wise fitting [16,17], the iterative regularized Gauss-Newton method (IRGNM) [18], simultaneous estimation of all parameters, L 1 regularization and the fast iterative shrinkage-thresholding algorithm (FISTA) [19]. In that case, despite application of external libraries and GPU acceleration, the fastest offline calculations are still performed in 10-20 min and even more.…”

“…In multi-channel systems the number of coils may be limited in the preprocessing step. The coil compression can be applied by singular value decomposition [34] or by evaluation of virtual channels using a principle component analysis [18,19,33].…”