Magnetic Resonance Materials in Physics, Biology and Medicine最新文献

Deep-learning-based MR image reconstruction in settings where large fully sampled dataset collection is infeasible requires methods that effectively use both under-sampled and fully sampled datasets. This paper evaluates a weakly supervised, multi-coil, physics-guided approach to MR image reconstruction, leveraging both dataset types, to improve both the quality and robustness of reconstruction. A physics-guided end-to-end variational network (VarNet) is pretrained in a self-supervised manner using a 4 $\times$ under-sampled dataset following the self-supervised learning via data undersampling (SSDU) methodology. The pre-trained weights are transferred to another VarNet, which is fine-tuned using a smaller, fully sampled dataset by optimizing multi-scale structural similarity (MS-SSIM) loss in image space. The proposed methodology is compared with fully self-supervised and fully supervised training. Reconstruction quality improvements in SSIM, PSNR, and NRMSE when abundant training data is available (the high-data regime), and enhanced robustness when training data is scarce (the low-data regime) are demonstrated using weak supervision for knee and brain MR image reconstructions at 8 $\times$ and 10 $\times$ acceleration, respectively. Multi-coil physics-guided MR image reconstruction using both under-sampled and fully sampled datasets is achievable with transfer learning and fine-tuning. This methodology can provide improved reconstruction quality in the high-data regime and improved robustness in the low-data regime at high acceleration rates.

Objective: This work presents an automated quality control (QC) system within quantitative MRI (qMRI) workflows. By leveraging the ISMRM/NIST quantitative MRI system phantom, we establish an open-source pipeline for rapid, repeatable, and accurate validation and stability tracking of sequence quantification performance across diverse clinical settings.

Materials and methods: A microservice-based QC system for automated vial segmentation from quantitative maps was developed and tested across various MRF acquisition and protocol designs, with reports generated and returned to the scanner in real time.

Results: The system demonstrated consistent and repeatable value segmentation and reporting, successfully extracted all 252 T1 and T2 vial samples tested. Values extracted from the same sequence were found to be repeatable with 0.09% ± 1.23% and - 0.26% ± 2.68% intersession error, respectively.

Discussion: By providing real-time quantification performance assessment, this easily deployable automated QC approach streamlines sequence validation and long-term performance monitoring, vital for the broader acceptance of qMRI as a standard component of clinical protocols.

Objective: Glioblastoma multiforme is a highly aggressive form of brain cancer, and early diagnosis plays a pivotal role in improving patient survival rates. In this regard, molecular magnetic resonance imaging has emerged as a promising imaging modality due to its exceptional sensitivity to minute tissue changes and the ability to penetrate deep into the brain. This study aimed to assess the efficacy of a novel contrast agent in detecting gliomas during MRI scans.

Materials and methods: The contrast agent utilized modified chitosan coating on manganese oxide nanoparticles. The modification included adding methotrexate and 5-aminolevulinic acid (MnO₂/CS@5-ALA-MTX) to target cells with overexpressed folate receptors and breaking down excess hydrogen peroxide in tumor tissue, resulting in enhanced signal intensity in T₁-weighted MR images but diminished signal intensity in T₂*-weighted MR images.

Results: The nanosystem was characterized and evaluated in MR imaging, safety, and ability to target cells both in vivo and in vitro. MTX-free nanoparticles (MnO₂/CS@5-ALA NPs) had no obvious cytotoxicity on cell lines U87MG and NIH3T3 after 24/48 h at a concentration of up to 160 µgr/mL (cell viability more than 80%). In this system, methotrexate enables tumor targeting and the MnO₂/5-ALA improves T₁-T₂^*-weighted MRI. In addition, MRI scans of mice with M109 carcinoma indicated significant tumor uptake and NP capacity to improve the positive contrast effect.

Conclusion: This developed MnO₂/CS@5-ALA-MTX nanoparticle system may exhibit great potential in the accurate diagnosis of folate receptor over-expressing cancers such as glioblastoma.

Objective: To prepare and analyze soy-lecithin-agar gels for non-toxic relaxometry phantoms with tissue-like relaxation times at 3T.

Methods: Phantoms mimicking the relaxation times of various tissues (gray and white matter, kidney cortex and medulla, spleen, muscle, liver) were built and tested with a clinical 3T whole-body MR scanner. Simple equations were derived to calculate the appropriate concentrations of soy lecithin and agar in aqueous solutions to achieve the desired relaxation times. Phantoms were tested for correspondence between measurements and calculated T1 and T2 values, reproducibility, spatial homogeneity, and temporal stability. T1 and T2 mapping techniques and a 3D T1-weighted sequence with high spatial resolution were applied.

Results: Except for the liver relaxation phantom, all phantoms were successfully and reproducibly produced. Good agreement was found between the targeted and measured relaxation times. The percentage deviations from the targeted relaxation times were less than 3% for T1 and less than 6.5% for T2. In addition, the phantoms were homogeneous and had little to no air bubbles. However, the phantoms were unstable over time: after a storage period of 4 weeks, mold growth and also changes in relaxation times were detected in almost all phantoms.

Conclusion: Soy-lecithin-agar gels are a non-toxic material for the construction of relaxometry phantoms with tissue-like relaxation times. They are easy to prepare, inexpensive and allow independent adjustment of T1 and T2. However, there is still work to be done to improve the long-term stability of the phantoms.

Object: In a typical MR session, several contrasts are acquired. Due to the sequential nature of the data acquisition process, the patient may experience some discomfort at some point during the session, and start moving. Hence, it is quite common to have MR sessions where some contrasts are well-resolved, while other contrasts exhibit motion artifacts. Instead of repeating the scans that are corrupted by motion, we introduce a reference-guided retrospective motion correction scheme that takes advantage of the motion-free scans, based on a generalized rigid registration routine.

Materials and methods: We focus on various existing clinical 3D brain protocols at 1.5 Tesla MRI based on Cartesian sampling. Controlled experiments with three healthy volunteers and three levels of motion are performed.

Results: Radiological inspection confirms that the proposed method consistently ameliorates the corrupted scans. Furthermore, for the set of specific motion tests performed in this study, the quality indexes based on PSNR and SSIM shows only a modest decrease in correction quality as a function of motion complexity.

Discussion: While the results on controlled experiments are positive, future applications to patient data will ultimately clarify whether the proposed correction scheme satisfies the radiological requirements.

Objective: Many patients with long COVID experience neurological and psychological symptoms. Signal abnormalities on MR images in the corpus callosum have been reported. Knowledge about the metabolic profile in the splenium of the corpus callosum (CCS) may contribute to a better understanding of the pathophysiology of long COVID.

Materials and methods: Eighty-one subjects underwent proton MR spectroscopy examination. The metabolic concentrations of total N-acetylaspartate (NAA), choline-containing compounds (Cho), total creatine (Cr), myo-inositol (mI), and NAA/Cho in the CCS were statistically compared in the group of patients containing 58 subjects with positive IgG COVID-19 antibodies or positive SARS-CoV-2 qPCR test at least two months before the MR and the group of healthy controls containing 23 subjects with negative IgG antibodies.

Results: An age-dependent effect of SARS-CoV-2 on Cho concentrations in the CCS has been observed. Considering the subjective threshold of age = 40 years, older patients showed significantly increased Cho concentrations in the CCS than older healthy controls (p = 0.02). NAA, Cr, and mI were unchanged. All metabolite concentrations in the CCS of younger post-COVID-19 patients remained unaffected by SARS-CoV-2. Cho did not show any difference between symptomatic and asymptomatic patients (p = 0.91).

Discussion: Our results suggest that SARS-CoV-2 disproportionately increases Cho concentration in the CCS among older post-COVID-19 patients compared to younger ones. The observed changes in Cho may be related to the microstructural reorganization in the CCS also reported in diffusion measurements rather than increased membrane turnover. These changes do not seem to be related to neuropsychological problems of the post-COVID-19 patients. Further metabolic studies are recommended to confirm these observations.

Objectives: To evaluate a new motion correction method, named RT + NV Track, for upper abdominal DWI that combines the respiratory triggering (RT) method using a respiration sensor and the Navigator Track (NV Track) method using navigator echoes.

Materials and methods: To evaluate image quality acquired upper abdominal DWI and ADC images with RT, NV, and RT + NV Track in 10 healthy volunteers and 35 patients, signal-to-noise efficiency (SNR_efficiency) and the coefficient of variation (CV) of ADC values were measured. Five radiologists independently performed qualitative image-analysis assessments.

Results: RT + NV Track showed significantly higher SNR_efficiency than RT and NV (14.01 ± 4.86 vs 12.05 ± 4.65, 10.05 ± 3.18; p < 0.001, p < 0.001). RT + NV Track was superior to RT and equal or better quality than NV in CV and visual evaluation of ADC values (0.033 ± 0.018 vs 0.080 ± 0.042, 0.057 ± 0.034; p < 0.001, p < 0.001). RT + NV Track tends to acquire only expiratory data rather than NV, even in patients with relatively rapid breathing, and can correct for respiratory depth variations, a weakness of RT, thus minimizing image quality degradation.

Conclusion: The RT + NV Track method is an efficient imaging method that combines the advantages of both RT and NV methods in upper abdominal DWI, providing stably good images in a short scan time.

Objective: Diffusional kurtosis imaging (DKI) extends diffusion tensor imaging (DTI), characterizing non-Gaussian diffusion effects but requires longer acquisition times. To ensure the robustness of DKI parameters, data acquisition ordering should be optimized allowing for scan interruptions or shortening. Three methodologies were used to examine how reduced diffusion MRI scans impact DKI histogram-metrics: 1) the electrostatic repulsion model (Opt_EEM); 2) spherical codes (Opt_SC); 3) random (Random_TRUNC).

Materials and methods: Pre-acquired diffusion multi-shell data from 14 female healthy volunteers (29±5 years) were used to generate reordered data. For each strategy, subsets containing different amounts of the full dataset were generated. The subsampling effects were assessed on histogram-based DKI metrics from tract-based spatial statistics (TBSS) skeletonized maps. To evaluate each subsampling method on simulated data at different SNRs and the influence of subsampling on in vivo data, we used a 3-way and 2-way repeated measures ANOVA, respectively.

Results: Simulations showed that subsampling had different effects depending on DKI parameter, with fractional anisotropy the most stable (up to 5% error) and radial kurtosis the least stable (up to 26% error). Random_TRUNC performed the worst while the others showed comparable results. Furthermore, the impact of subsampling varied across distinct histogram characteristics, the peak value the least affected (Opt_EEM: up to 5% error; Opt_SC: up to 7% error) and peak height (Opt_EEM: up to 8% error; Opt_SC: up to 11% error) the most affected.

Conclusion: The impact of truncation depends on specific histogram-based DKI metrics. The use of a strategy for optimizing the acquisition order is advisable to improve DKI robustness to exam interruptions.

Background: Magnetic Resonance Imaging (MRI) is a highly demanded medical imaging system due to high resolution, large volumetric coverage, and ability to capture the dynamic and functional information of body organs e.g. cardiac MRI is employed to assess cardiac structure and evaluate blood flow dynamics through the cardiac valves. Long scan time is the main drawback of MRI, which makes it difficult for the patients to remain still during the scanning process.

Objective: By collecting fewer measurements, MRI scan time can be shortened, but this undersampling causes aliasing artifacts in the reconstructed images. Advanced image reconstruction algorithms have been used in literature to overcome these undersampling artifacts. These algorithms are computationally expensive and require a long time for reconstruction which makes them infeasible for real-time clinical applications e.g. cardiac MRI. However, exploiting the inherent parallelism in these algorithms can help to reduce their computation time.

Methods: Low-rank plus sparse (L+S) matrix decomposition model is a technique used in literature to reconstruct the highly undersampled dynamic MRI (dMRI) data at the expense of long reconstruction time. In this paper, Compressed Singular Value Decomposition (cSVD) model is used in L+S decomposition model (instead of conventional SVD) to reduce the reconstruction time. The results provide improved quality of the reconstructed images. Furthermore, it has been observed that cSVD and other parts of the L+S model possess highly parallel operations; therefore, a customized GPU based parallel architecture of the modified L+S model has been presented to further reduce the reconstruction time.

Results: Four cardiac MRI datasets (three different cardiac perfusion acquired from different patients and one cardiac cine data), each with different acceleration factors of 2, 6 and 8 are used for experiments in this paper. Experimental results demonstrate that using the proposed parallel architecture for the reconstruction of cardiac perfusion data provides a speed-up factor up to 19.15× (with memory latency) and 70.55× (without memory latency) in comparison to the conventional CPU reconstruction with no compromise on image quality.

Conclusion: The proposed method is well-suited for real-time clinical applications, offering a substantial reduction in reconstruction time.