An empirical study of the maximum degree of undersampling in compressed sensing for T2*-weighted MRI

Optimized and accelerated 19 F-MRI of inhaled perfluoropropane to assess regional pulmonary ventilation

PURPOSE

To accelerate ¹⁹ F-MR imaging of inhaled perfluoropropane using compressed sensing methods, and to optimize critical scan acquisition parameters for assessment of lung ventilation properties.

METHODS

Simulations were performed to determine optimal acquisition parameters for maximal perfluoropropane signal-to-noise ratio (SNR) in human lungs for a spoiled gradient echo sequence. Optimized parameters were subsequently employed for ¹⁹ F-MRI of inhaled perfluoropropane in a cohort of 11 healthy participants using a 3.0 T scanner. The impact of 1.8×, 2.4×, and 3.0× undersampling ratios on ¹⁹ F-MRI acquisitions was evaluated, using both retrospective and prospective compressed sensing methods.

RESULTS

3D spoiled gradient echo ¹⁹ F-MR ventilation images were acquired at 1-cm isotropic resolution within a single breath hold. Mean SNR was 11.7 ± 4.1 for scans acquired within a single breath hold (duration = 18 s). Acquisition of ¹⁹ F-MRI scans at shorter scan durations (4.5 s) was also demonstrated as feasible. Application of both retrospective (n = 8) and prospective (n = 3) compressed sensing methods demonstrated that 1.8× acceleration had negligible impact on qualitative image appearance, with no statistically significant change in measured lung ventilated volume. Acceleration factors of 2.4× and 3.0× resulted in increasing differences between fully sampled and undersampled datasets.

CONCLUSION

This study demonstrates methods for determining optimal acquisition parameters for ¹⁹ F-MRI of inhaled perfluoropropane and shows significant reduction in scan acquisition times (and thus participant breath hold duration) by use of compressed sensing.

A Reweighted Symmetric Smoothed Function Approximating L0-Norm Regularized Sparse Reconstruction Method

Sparse-signal recovery in noisy conditions is a problem that can be solved with current compressive-sensing (CS) technology. Although current algorithms based on L 1 regularization can solve this problem, the L 1 regularization mechanism cannot promote signal sparsity under noisy conditions, resulting in low recovery accuracy. Based on this, we propose a regularized reweighted composite trigonometric smoothed L 0 -norm minimization (RRCTSL0) algorithm in this paper. The main contributions of this paper are as follows: (1) a new smoothed symmetric composite trigonometric (CT) function is proposed to fit the L 0 -norm; (2) a new reweighted function is proposed; and (3) a new L 0 regularization objective function framework is constructed based on the idea of T i k h o n o v regularization. In the new objective function framework, Contributions (1) and (2) are combined as sparsity regularization terms, and errors as deviation terms. Furthermore, the conjugate-gradient (CG) method is used to optimize the objective function, so as to achieve accurate recovery of sparse signal and image under noisy conditions. The numerical experiments on both the simulated and real data verify that the proposed algorithm is superior to other state-of-the-art algorithms, and achieves advanced performance under noisy conditions.