MRI深度学习图像重建技术在肌骨系统疾病诊断的应用进展

doi:10.16150/j.1671-2870.2024.02.003

摘要/Abstract

摘要：

深度学习图像重建（deep learning-based reconstruction，DLR）技术是目前MRI图像重建领域最为前沿的技术进展之一。相对于常规MRI图像重建技术而言，DLR技术重新定义了MRI的信噪比、空间分辨率和扫描时间之间新的边界，其突出的技术优势是有效去除图像噪声及伪影，大幅缩短扫描时间，且在提高病灶的检出率和定性准确率方面也具有潜在优势。随着算法的不断优化和模型泛化性的提升，DLR目前已被广泛应用于神经系统、肌骨系统及心脏等多部位的MRI检查，其适用的扫描序列及临床应用场景也在不断拓展。DLR技术在维持原有空间分辨率条件下，通过减少信号采集次数联合增加并行采集加速因子，将成像时间缩短50%以上，实现肌骨系统快速成像，且所获得图像质量明显优于传统重建图像。目前，DLR在膝关节、肩关节、手腕关节及脊柱等肌骨系统的MRI检查中被广泛应用，并证实了其在缩短成像时间、提升图像信噪比和提高分辨率方面具有卓越表现。

关键词: 磁共振成像, 深度学习图像重建, 肌骨系统

Abstract:

Deep learning based reconstruction (DLR) technology is currently one of the most cutting-edge technological advancements in the field of MRI image reconstruction. Compared to conventional MRI image reconstruction techniques, DLR technology redefines a new boundary between signal-to-noise ratio, spatial resolution, and scanning time on MRI. Its outstanding technical advantage is the effective removal of image noise and artifacts, significantly reducing scanning time, and also has potential advantages in improving the detection rate and accuracy qualitative diagnosis of lesions. With the continuous optimization of algorithms and the improvement of model generalization, DLR has been widely used in MRI examinations for multiple parts, such as the nervous system, musculoskeletal system, and heart. Its applicable scanning sequences and clinical application scenarios are also constantly expanding. DLR technology, while maintaining the original spatial resolution, reduces the number of signal acquisition times and increases the parallel acquisition acceleration factor to shorten the imaging time by more than 50%, achieving rapid imaging of the musculoskeletal system, and obtaining significantly better image quality than traditional reconstructed images. Currently, DLR is widely used in MRI exa-minations of musculoskeletal systems, such as the knee, shoulder, wrist, and spine, and has demonstrated its outstanding performance in shortening imaging time, improving image signal-to-noise ratio and resolution.

Key words: Magnetic resonance imaging, Deep learning-based reconstruction, Musculoskeletal system

中图分类号:

R681

查云飞, 武夏夏. MRI深度学习图像重建技术在肌骨系统疾病诊断的应用进展[J]. 诊断学理论与实践, 2024, 23(02): 114-118.

ZHA Yunfei, WU Xiaxia. Application and research progress of MRI deep learning image reconstruction technology in clinical diagnosis of musculoskeletal system diseases[J]. Journal of Diagnostics Concepts & Practice, 2024, 23(02): 114-118.

参考文献 25

[1]	CHAUDHARI A S, STEVENS K J, SVEINSSON B, et al. Combined 5-minute double-echo in steady-state with separated echoes and 2-minute proton-density-weighted 2D FSE sequence for comprehensive whole-joint knee MRI assessment[J]. J Magn Reson Imaging, 2019, 49(7):e183-e194.
[2]	严福华. 深度学习图像重建算法的临床应用和发展前景[J]. 中华放射学杂志, 2022, 56(11):1165-1167.
	YAN F H. The clinical application and development prospect of deep learning reconstruction algorithm[J]. Chin J Radiol, 2022, 56(11):1165-1167.
[3]	FRITZ J, GUGGENBERGER R, DEL GRANDE F. Rapid musculoskeletal MRI in 2021: clinical application of advanced accelerated techniques[J]. AJR, 2021, 216(3):718-733.
[4]	HAHN S, YI J, LEE HJ, et al. Comparison of deep lear-ning-based reconstruction of PROPELLER Shoulder MRI with conventional reconstruction[J]. Skeletal Radiol, 2023, 52(8):1545-1555.
[5]	YANG R, ZOU Y, LIU W V, et al. High-resolution single-shot fast spin-echo mr imaging with deep learning reconstruction algorithm can improve repeatability and reproducibility of follicle counting[J]. J Clin Med, 2023, 12(9):3234.
[6]	HILBERT T, SUMPF T J, WEILAND E, et al. Accele-rated T2 mapping combining parallel MRI and model-based reconstruction: GRAPPATINI[J]. J Magn Reson Imaging, 2018, 48(2):359-368.
[7]	CHAUDHARI A S, GRISSOM M J, FANG Z, et al. Diagnostic accuracy of quantitative multicontrast 5-minute knee MRI using prospective artificial intelligence image quality enhancement[J]. AJR, 2021, 216(6):1614-1625.
[8]	WANG S S, XIAO T H, LIU Q G, et al. Deep learning for fast MR imaging: a review for learning reconstruction from incomplete k-space data[J]. Biomed Signal Process Control, 2021, 68(5):102579.
[9]	MANCO L, MAFFEI N, STROLIN S, et al. Basic of machine learning and deep learning in imaging for medical physicists[J]. Phys Med, 2021, 83:194-205. doi: 10.1016/j.ejmp.2021.03.026 pmid: 33826964
[10]	PETERS R D, HARRIS H, LAWSON S. The clinical be-nefits of AIRTM Recon DL for MR image reconstruction[EB/OL]. [2022-11-07]. https://www.gehealthcare.com/ensg/-/jssmedia/c943df5927a049bb9ac95a9f0349ad8c.pdf.
[11]	LEBEL R M. Performance characterization of a novel deep learning-based MR image reconstruction pipeline[EB/OL]. [2022-11-07]. https://arxiv.org/ftp/arxiv/papers/2008/2008.06559.pdf.
[12]	HERRMANN J, KOERZDOERFER G, NICKEL D, et al. Feasibility and implementation of a deep learning mr reconstruction for TSE sequences in musculoskeletal imaging[J]. Diagnostics(Basel), 2021, 11(8):1484.
[13]	RECHT M P, ZBONTAR J, SODICKSON D K, et al. Usi-ng deep learning to accelerate knee MRI at 3 T: results of an interchangeability study[J]. AJR, 2020, 215(6):1421-1429.
[14]	JOHNSON P M, LIN D J, ZBONTAR J, et al. Deep lear-ning reconstruction enables prospectively accelerated clinical knee MRI[J]. Radiology, 2023, 307(2):e220425.
[15]	武夏夏, 陆雪芳, 刘昌盛, 等. 深度学习图像重建算法在膝关节加速MRI中的临床应用研究[J]. 磁共振成像, 2023, 14(5):53-59.
	WU X X, LU X F, LIU C S, et al. Clinical feasibility of 2D FSE sequences of the knee MRI protocol using deep-learning image reconstruction[J]. Chin J Magn Reson Ima-ging, 2023, 14(5):53-59.
[16]	HERRMANN J, KELLER G, GASSENMAIER S, et al. Feasibility of an accelerated 2D-multi-contrast knee MRI protocol using deep-learning image reconstruction: a prospective intraindividual comparison with a standard MRI protocol[J]. Eur Radiol, 2022, 32(9):6215-6229. doi: 10.1007/s00330-022-08753-z pmid: 35389046
[17]	CHANG P D, CHOW D S. Revolutionizing Shoulder MRI: Accelerated Imaging with Deep Learning Reconstruction[J]. Radiology, 2024, 310(1):e233301.
[18]	HAHN S, YI J, LEE H J, et al. Image quality and diagnostic performance of accelerated shoulder MRI with deep learning-based reconstruction[J]. AJR, 2022, 218(3):506-516.
[19]	KANIEWSKA M, DEININGER-CZERMAK E, GETZMANN J M, et al. Application of deep learning-based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time[J]. Eur Radiol, 2023, 33(3):1513-1525.
[20]	XIE Y, TAO H, LI X, et al. Prospective comparison of standard and deep learning-reconstructed turbo spin-echo MRI of the shoulder[J]. Radiology, 2024, 310(1):e231405.
[21]	HERRMANN J, GASSENMAIER S, KELLER G, et al. Deep Learning MRI reconstruction for accelerating turbo spin echo hand and wrist imaging: A comparison of image quality, visualization of anatomy, and detection of common pathologies with standard imaging[J]. Acad Radiol, 2023, 30(11):2606-2615.
[22]	YASAKA K, TANISHIMA T, OHTAKE Y, et al. Deep learning reconstruction for the evaluation of neuroforaminal stenosis using 1.5T cervical spine MRI: comparison with 3T MRI without deep learning reconstruction[J]. Neuroradiology, 2022, 64(10):2077-2083. doi: 10.1007/s00234-022-03024-6 pmid: 35918450
[23]	JARDON M, TAN E T, CHAZEN J L, et al. Deep-learni-ng-reconstructed high-resolution 3D cervical spine MRI for foraminal stenosis evaluation[J]. Skeletal Radiol, 2023, 52(4):725-732.
[24]	SUN S, TAN E T, MINTZ D N, et al. Evaluation of deep learning reconstructed high-resolution 3D lumbar spine MRI[J]. Eur Radiol, 2022, 32(9):6167-6177.
[25]	ALMANSOUR H, HERRMANN J, GASSENMAIER S, et al. Deep learning reconstruction for accelerated spine MRI: prospective analysis of interchangeability[J]. Radio-logy, 2023, 306(3):e212922.