Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Jun 16;20(6):e0325783.
doi: 10.1371/journal.pone.0325783. eCollection 2025.

Shortening 7T MP2RAGE acquisition with compressed sensing: Evaluating quantitative accuracy and structural consistency

Ikuhiro Kida  1
Affiliations

Shortening 7T MP2RAGE acquisition with compressed sensing: Evaluating quantitative accuracy and structural consistency

Ikuhiro Kida. PLoS One. .

Abstract

The aim of this study was to systematically evaluate the impact of compressed sensing (CS) on acquisition time, image quality, T1 mapping accuracy, and segmentation consistency in magnetization-prepared 2 rapid acquisition gradient echo (MP2RAGE) at ultra-high fields (UHF). MP2RAGE sequences were acquired using the CS and parallel imaging (PI) technique, i.e., generalized autocalibrating partially parallel acquisitions (GRAPPA), with varying undersampling factors and samples per repetition time (TR). The acquisition time, quantitative accuracy of T1 mapping, and segmentation consistency across regions of interest (ROIs) were assessed. CS-MP2RAGE achieved a 61% reduction in acquisition time (< 3 min) compared with PI-MP2RAGE and maintained comparable image quality, segmentation accuracy, and T1-mapping fidelity. Higher undersampling factors effectively reduced scan duration but introduced segmentation volume mismatches of up to 20% and increased T1 values, despite the images appearing similar to PI-MP2RAGE. Reducing the number of samples per TR enhanced image quality, allowing for higher undersampling factors without a significant loss of fidelity, a finding consistent with previous studies. However, excessively low sampling densities destabilized reconstruction in complex ROIs. Our findings demonstrate that CS-MP2RAGE significantly reduces scan time while maintaining high image quality and quantitative accuracy, making it a viable alternative to GRAPPA in UHF applications. The interplay between undersampling factors and samples per TR is crucial for optimizing scan efficiency. Future studies should explore its application in clinical and research settings.

PubMed Disclaimer

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1
Fig 1. Comparison of T1-weighted images between PI-MP2RAGE and CS-MP2RAGE across all undersampling factors and samples per TR.
Axial plane images from a representative subject are shown. The labels at the bottom of each image indicate the undersampling factors and the corresponding samples per TR. Red contour lines indicate gray matter segmentation in the left hemisphere.
Fig 2
Fig 2. Comparison of voxel counts and segmentation overlap across brain regions between CS-MP2RAGE and PI-MP2RAGE.
(A, B) Changes in voxel counts within each region of interest (ROI) based on the Desikan–Killiany atlas in FreeSurfer, comparing PI-MP2RAGE and CS-MP2RAGE. (A) Results for CS-MP2RAGE undersampling factors 3.8, 4.4, and 8.0, with corresponding samples per TR of 250, 252, and 253, respectively. (B) Results for CS-MP2RAGE undersampling factors 3.8, 8.0, 8.0, and 10.0, with samples per TR of 196, 192, 126, and 126, respectively. (C, D) DSCs for each ROI, representing the degree of segmentation overlap between PI-MP2RAGE and CS-MP2RAGE. (C) Results for CS-MP2RAGE undersampling factors 3.8, 4.4, and 8.0, with samples per TR of 250, 252, and 253, respectively. (D) Results for CS-MP2RAGE undersampling factors 3.8, 8.0, 8.0, and 10.0, with samples per TR of 196, 192, 126, and 126, respectively. In (A), the colors of the ROI names indicate the number of CS-MP2RAGE conditions that show statistically significant differences compared to PI-MP2RAGE: red for three, yellow for two, and blue for one.
Fig 3
Fig 3. Contrast ratio between gray matter and white matter in PI-MP2RAGE and CS-MP2RAGE.
Results of CS-MP2RAGE are shown for undersampling factors of 3.8, 4.4, 8.0, and 10.0, with corresponding samples per TR of 3.8 (250), 4.4 (252), 8.0 (253), 3.8 (196), 8.0 (192), 8.0 (126), and 10.0 (126).
Fig 4
Fig 4. Comparison of T1 values across brain regions between CS-MP2RAGE and PI-MP2RAGE.
(A, B) T1 values within each region of interest (ROI) based on the Desikan–Killiany atlas in FreeSurfer. (A) Results for PI-MP2RAGE and CS-MP2RAGE undersampling factors 3.8, 4.4, and 8.0, with corresponding samples per TR of 250, 252, and 253, respectively. (B) Results for CS-MP2RAGE undersampling factors 3.8, 8.0, 8.0, and 10.0, with samples per TR of 196, 192, 126, and 126, respectively. (C, D) Change in T1 within each ROI, comparing PI-MP2RAGE and CS-MP2RAGE. (C) Results for CS-MP2RAGE undersampling factors 3.8, 4.4, and 8.0, with samples per TR of 250, 252, and 253, respectively. (D) Results for CS-MP2RAGE undersampling factors 3.8, 8.0, 8.0, and 10.0, with samples per TR of 196, 192, 126, and 126, respectively. In (C) and (D), the colors of the ROI names indicate the number of CS-MP2RAGE conditions that show statistically significant differences compared to PI-MP2RAGE: gray for four, red for three, yellow for two, and blue for one.
Fig 5
Fig 5. T1 distributions in PI-MP2RAGE and CS-MP2RAGE across the whole gray matter.
T1 mean values were obtained in the whole gray matter over all subjects.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Marques JP, Kober T, Krueger G, van der Zwaag W, Van de Moortele P-F, Gruetter R. MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field. Neuroimage. 2010;49(2):1271–81. doi: 10.1016/j.neuroimage.2009.10.002 - DOI - PubMed
    1. Choi U-S, Kawaguchi H, Kida I. Cerebral artery segmentation based on magnetization-prepared two rapid acquisition gradient echo multi-contrast images in 7 Tesla magnetic resonance imaging. Neuroimage. 2020;222:117259. doi: 10.1016/j.neuroimage.2020.117259 - DOI - PubMed
    1. Choi U-S, Kawaguchi H, Matsuoka Y, Kober T, Kida I. Brain tissue segmentation based on MP2RAGE multi-contrast images in 7 T MRI. PLoS One. 2019;14(2):e0210803. doi: 10.1371/journal.pone.0210803 - DOI - PMC - PubMed
    1. Kido T, Nakamura M, Watanabe K, Schmidt M, Forman C, et al. Compressed sensing real-time cine cardiovascular magnetic resonance: accurate assessment of left ventricular function in a single-breath-hold. J Cardiovasc Magn Reson. 2016; 18(1): 50. doi: 10.1186/s12968-016-0271-0 . - DOI - PMC - PubMed
    1. Fushimi Y, Fujimoto K, Okada T, Yamamoto A, Tanaka T, Kikuchi T, et al. Compressed Sensing 3-Dimensional Time-of-Flight Magnetic Resonance Angiography for Cerebral Aneurysms: Optimization and Evaluation. Invest Radiol. 2016;51(4):228–35. doi: 10.1097/RLI.0000000000000226 - DOI - PubMed