Zero-DeepSub: Zero-shot deep subspace reconstruction for rapid multiparametric quantitative MRI using 3D-QALAS.

PURPOSE: To develop and evaluate methods for (1) reconstructing 3D-quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) time-series images using a low-rank subspace method, which enables accurate and rapid T1 and T2 mapping, and (2) improving the fidelity of subspace QALAS by combining scan-specific deep-learning-based reconstruction and subspace modeling. THEORY AND METHODS: A low-rank subspace method for 3D-QALAS (i.e., subspace QALAS) and zero-shot deep-learning subspace method (i.e., Zero-DeepSub) were proposed for rapid and high fidelity T1 and T2 mapping and time-resolved imaging using 3D-QALAS. Using an ISMRM/NIST system phantom, the accuracy and reproducibility of the T1 and T2 maps estimated using the proposed methods were evaluated by comparing them with reference techniques. The reconstruction performance of the proposed subspace QALAS using Zero-DeepSub was evaluated in vivo and compared with conventional QALAS at high reduction factors of up to nine-fold. RESULTS: Phantom experiments showed that subspace QALAS had good linearity with respect to the reference methods while reducing biases and improving precision compared to conventional QALAS, especially for T2 maps. Moreover, in vivo results demonstrated that subspace QALAS had better g-factor maps and could reduce voxel blurring, noise, and artifacts compared to conventional QALAS and showed robust performance at up to nine-fold acceleration with Zero-DeepSub, which enabled whole-brain T1, T2, and PD mapping at 1 mm isotropic resolution within 2 min of scan time. CONCLUSION: The proposed subspace QALAS along with Zero-DeepSub enabled high fidelity and rapid whole-brain multiparametric quantification and time-resolved imaging.

SSL-QALAS: Self-Supervised Learning for rapid multiparameter estimation in quantitative MRI using 3D-QALAS.

PURPOSE: To develop and evaluate a method for rapid estimation of multiparametric T1 , T2 , proton density, and inversion efficiency maps from 3D-quantification using an interleaved Look-Locker acquisition sequence with T2 preparation pulse (3D-QALAS) measurements using self-supervised learning (SSL) without the need for an external dictionary. METHODS: An SSL-based QALAS mapping method (SSL-QALAS) was developed for rapid and dictionary-free estimation of multiparametric maps from 3D-QALAS measurements. The accuracy of the reconstructed quantitative maps using dictionary matching and SSL-QALAS was evaluated by comparing the estimated T1 and T2 values with those obtained from the reference methods on an International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology phantom. The SSL-QALAS and the dictionary-matching methods were also compared in vivo, and generalizability was evaluated by comparing the scan-specific, pre-trained, and transfer learning models. RESULTS: Phantom experiments showed that both the dictionary-matching and SSL-QALAS methods produced T1 and T2 estimates that had a strong linear agreement with the reference values in the International Society for Magnetic Resonance in Medicine/National Institute of Standards and Technology phantom. Further, SSL-QALAS showed similar performance with dictionary matching in reconstructing the T1 , T2 , proton density, and inversion efficiency maps on in vivo data. Rapid reconstruction of multiparametric maps was enabled by inferring the data using a pre-trained SSL-QALAS model within 10 s. Fast scan-specific tuning was also demonstrated by fine-tuning the pre-trained model with the target subject's data within 15 min. CONCLUSION: The proposed SSL-QALAS method enabled rapid reconstruction of multiparametric maps from 3D-QALAS measurements without an external dictionary or labeled ground-truth training data.

3D Synthetic Brain MRI with Compressed Sensing: Can It Be a Promising Way Forward for Daily Neuroimaging?

BACKGROUND: Synthetic MRI can provide multiple contrast-weighted brain images with high resolution from a single scan via a 3D sequence using an interleaved Look-Locker acquisition sequence with a T2 preparation pulse (3D-QALAS). OBJECTIVE: This study aimed to assess the diagnostic image quality of 3D synthetic MRI using compressed sensing (CS) in clinical practice. METHODS: We retrospectively reviewed the imaging data of 47 patients who underwent brain MRI, including 3D synthetic MRI using CS in a single session, between December 2020 and February 2021. Two neuroradiologists independently evaluated the overall image quality, anatomic demarcation, and artifacts for synthetic 3D T1-weighted, T2-weighted, FLAIR, phase-sensitive inversion recovery (PSIR), and double inversion recovery images, using a 5-point Likert scale. The interobserver agreement between the two readers was assessed using percent agreement and weighted κ statistics. RESULTS: The overall image quality of 3D synthetic T1WI and PSIR was good to excellent, with easy or excellent anatomic demarcation and mild or no visible artifact. However, other 3D synthetic MRI-derived images showed insufficient image quality and anatomic demarcation with marked CSF pulsation artifacts. In particular, 3D synthetic FLAIR showed high-signal artifacts on the brain surface. CONCLUSION: 3D synthetic MRI, at its current status, cannot completely replace conventional brain MRI in daily clinical practice. However, 3D synthetic MRI can achieve scan-time reduction using CS and parallel imaging and may be useful for motion-prone or pediatric patients requiring 3D images where time-efficiency is important.

Clinical feasibility of 3D-QALAS - Single breath-hold 3D myocardial T1- and T2-mapping.

PURPOSE: To investigate the in-vivo precision and clinical feasibility of 3D-QALAS - a novel method for simultaneous three-dimensional myocardial T1- and T2-mapping. METHODS: Ten healthy subjects and 23 patients with different cardiac pathologies underwent cardiovascular 3T MRI examinations including 3D-QALAS, MOLLI and T2-GraSE acquisitions. Precision was investigated in the healthy subjects between independent scans, between dependent scans and as standard deviation of consecutive scans. Clinical feasibility of 3D-QALAS was investigated for native and contrast enhanced myocardium in patients. Data were analyzed using mean value and 95% confidence interval, Pearson correlation, Paired t-tests, intraclass correlation and Bland-Altman analysis. RESULTS: Average myocardial relaxation time values and SD from eight repeated acquisitions within the group of healthy subjects were 1178±18.5ms (1.6%) for T1 with 3D-QALAS, 52.7±1.2ms (2.3%) for T2 with 3D-QALAS, 1145±10.0ms (0.9%) for T1 with MOLLI and 49.2±0.8ms (1.6%) for T2 with GraSE. Myocardial T1 and T2 relaxation times obtained with 3D-QALAS correlated very well with reference methods; MOLLI for T1 (r=0.994) and T2-GraSE for T2 (r=0.818) in the 23 patients. Average native/post-contrast myocardial T1 values from the patients were 1166.2ms/411.8ms for 3D-QALAS and 1174.4ms/438.9ms for MOLLI. Average native myocardial T2 values from the patients were 53.2ms for 3D-QALAS and 54.4ms for T2-GraSE. CONCLUSIONS: Repeated independent and dependent scans together with the intra-scan repeatability, demonstrated all a very good precision for the 3D-QALAS method in healthy volunteers. This study shows that 3D T1 and T2 mapping in the left ventricle is feasible in one breath hold for patients with different cardiac pathologies using 3D-QALAS.