Magnetic Resonance in Medicine最新文献

Purpose: To investigate the impact of iron particle size on $R_2^{*}$ and fat fraction (FF) estimations for coexisting hepatic iron overload and steatosis condition using Monte Carlo simulations and phantoms.

Methods: Three iron particle sizes (0.38, 0.52, and 0.71 μm) were studied using simulations and phantoms. Virtual liver models mimicking in vivo spatial distribution of fat droplets and iron deposits were created, and MRI signals were synthesized using Monte Carlo simulations for FF 1%-30% and liver iron concentration (LIC) 1-20 mg/g. Seventy-five fat-iron phantoms with varying iron (0-8 μg/mL) and fat (0%-40%) concentrations and particle sizes were constructed. Three-way analysis of variance was used to assess the effect of iron particle size on $R_2^{*}$ and FF estimations.

Results: In simulations, estimated and true FF were in excellent agreement (slope: 0.93-1.09) for liver iron concentration ≤ 13 mg/g. For both simulations and phantoms, FF estimation bias increased as iron concentration increased and particle size decreased, with 0.71μm iron particle having the lowest bias (≤ 20%), and 0.52 μm and 0.38 μm iron particles producing higher bias (≥ 20%) for higher iron concentrations and lower FFs. Additionally, $R_2^{*}$ increased linearly with increasing iron concentration (r ≥ 0.87) and decreasing particle size. Iron particle size significantly influenced the estimated versus true FF (simulations: p = 0.04; phantoms: p = 0.03) and $R_2^{*}$ -iron concentration (simulations: p < 0.001; phantoms: p < 0.01) relationships. Heatmap demonstrated broader region with higher FF estimation bias as iron particle size decreased, especially at higher iron concentration.

Conclusion: $R_2^{*}$ and FF estimations are affected by iron particle size, with smaller particles leading to higher $R_2^{*}$ values and increased FF estimation bias.

Purpose: To measure and validate elevated succinate in brain during circulatory arrest in a piglet model of cardiopulmonary bypass.

Methods: Using data from an archive of 3T ¹H MR spectra acquired in previous in-magnet studies, dynamic plots of succinate, spectral simulations and difference spectra were generated for analysis and validation.

Results: Elevation of succinate during circulatory arrest was observed and validated. Fitting bias was evaluated as a function of the line-widths and signal-to-noise ratios of the archived data. Succinate increases were independent of bypass temperature. Succinate elevation was also not observed with antegrade cerebral perfusion.

Conclusion: Although spectrally overlapped and at sub-millimolar levels, elevated brain succinate can be reliably measured by dynamic MR spectroscopy at 3T. Noise dependent bias of the stronger overlapping signals did not impact the succinate measurement. Elevated succinate during circulatory arrest and its recovery after reperfusion was observed. This finding is consistent with earlier reports that correlate elevated succinate with ischemic-reperfusion injury.

Purpose: Human brain development during gestation is complex, as both structure and function are rapidly forming. Structural imaging methods using MRI are well developed to explore these changes, but functional imaging tools are lacking. Low-field MRI is a promising modality to bridge this gap. The longer intrinsic T₂* values at low field strengths increase the dynamic range and enable the quantification of individual brain regions with low T₂* values, such as deep gray matter. This study investigates regional brain T₂* quantification throughout the second half of gestation on low-field 0.55T MRI.

Methods: Dynamic multi-echo gradient-echo sequences were acquired in 135 cases at 0.55 T between 20 and 40 weeks' gestation. Automatic high-resolution reconstruction and segmentation tools were developed, resulting in T₂* values of seven individual anatomical brain structures for each subject. These regional brain T₂* values were analyzed throughout gestation.

Results: All regional fetal brain T₂* values decreased throughout gestation (p < 0.01). Each anatomical brain structure had varying ranges and decay rates, with the cerebellum and white matter displaying the highest (nonfluid structure) values, with the maximum values between 350 and 400 ms at about 20 weeks. The brainstem and deep gray matter had the lowest range of T₂* values, reaching values of 250 ms early in gestation. The matched volumetric assessment of the different structures demonstrated expected growth, matching current literature.

Conclusion: Low-field MRI allows for a detailed, regional T₂* analysis of the fetal brain, with more inclusive norms to be developed due to its wider bore.

Purpose: The purpose of this study was to investigate microstructural changes in the aging adult prostate by comparing the effects of varying diffusion times using diffusion MRI, and to provide an age-related benchmark for future prostate cancer studies.

Methods: The prostates of normal male volunteers (n = 70, 19-69 years) were scanned at 3 T with an oscillating gradient spin echo (OGSE: 6 ms), pulsed gradient spin echo (PGSE: 40 ms) and pulsed gradient stimulated echo (PGSTE: 100 ms), and anatomical T₂-weighted image. Volume and mean diffusivity (MD) were measured in the peripheral (PZ) and transition zones (TZ), which were assessed versus age.

Results: PZ and TZ showed quadratic age trajectories for all diffusion scans, with MD decreasing from 19 years to a minimum ˜30-40 years followed by a greater increase at older ages. Short (OGSE) and medium (PGSE) diffusion time MD had similar age trajectories, whereas long diffusion time (PGSTE) MD was significantly lower, particularly in PZ (22%). MD difference (∆MD) of OGSE-PGSTE and PGSE-PGSTE showed significant positive linear correlations with age for both PZ (larger slope) and TZ, resulting in ˜3.3x (PZ) and 1.8x (TZ) greater ∆MD from 19 to 69 years. MD and ∆MD versus age relationships differed from volume, which conversely had greater proportional growth in TZ than PZ.

Conclusion: The diffusion time effects suggest age-related microstructural changes consistent with development of persistently larger cell dimensions mainly in the prostate peripheral zone over the adult lifespan. This normative data can be used for comparison to prostate cancer factoring in age.

Purpose: To implement a low-rank and subspace model-based reconstruction for 3D deuterium metabolic imaging (DMI) and compare its performance against Fourier transform-based (FFT) reconstruction in terms of spectral fitting reliability.

Methods: Both reconstruction methods were applied on simulated and experimental DMI data. Numerical simulations were performed to evaluate the effect of increasing acceleration factors. The impact on spectral fitting results, SNR, and the overall normalized root mean square error (NRMSE) compared to ground-truth data were calculated. A comparative analysis was performed on DMI data acquired from the human liver, including both natural abundance and post-deuterated glucose intake data at 7 T.

Results: Simulation showed the Cramer-Rao lower bound [%] of water, glucose, sum of glutamate and glutamine (Glx), and lipid signals for the low-rank and subspace model-based reconstruction at R = 1.0 was 12.4, 14.7, 17.3, and 11.0 times lower than FFT. At R = 1.1, NRMSE was 1.4%, 1.3%, 0.8%, and 4.2% lower for the water, glucose, Glx, and lipid, respectively, compared to FFT. However, the NRMSE of the Glx and lipid increased by 0.4% and 3.2% at R = 1.3. For the in vivo DMI experiment, SNR was 2.5-3.0 times higher compared to FFT. The fitted amplitude of water and glucose peaks showed Cramer-Rao lower bound [%] values that were approximately 2.3 times lower than FFT.

Conclusion: Simulations and in vivo experiments on the human liver demonstrate that low-rank and subspace model-based reconstruction with undersampled data mitigates noise and enhances spectral fitting quality.