Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Journal of Artificial Intelligence Practice
  4. Vol 9, Issue 1, 2026

Accelerated Cardiac MRI T1 Mapping Reconstruction Using Gated Linear Attention Network

Download as PDF

DOI: 10.23977/jaip.2026.090111 | Downloads: 1 | Views: 44

Author(s)

Chenqi Liu 1

Affiliation(s)

1 School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China

Corresponding Author

Chenqi Liu

ABSTRACT

Quantitative cardiac magnetic resonance (CMR) imaging often relies on high undersampling techniques to reduce scan time and suppress motion artifacts. However, extremely high acceleration rates inevitably lead to severe image aliasing and the loss of anatomical details. Existing reconstruction methods based on deep unrolling models face a dual challenge when processing dynamic sequences: severe error accumulation in early iterations and massive computational overhead for high-dimensional spatio-temporal feature extraction. To address these issues, this paper proposes an efficient and lightweight CMR reconstruction network (EGLAVarNet). First, to tackle the "cold start" problem in early iterations, a Content-Aware Structured Initialization (CASI) module is designed. This module leverages explicit low-rank physical priors to dynamically synthesize the initial state of the memory matrix, effectively suppressing the interference of high-rank random artifacts. Second, a Gated Linear Attention (GLA) unit is introduced in the core denoising stage. By utilizing the associative property of matrix multiplication to extract global structural features, this module reduces the time complexity of attention computation to a linear scale of O(N) without requiring spatial downsampling. Experimental results on the public CMRxRecon dataset demonstrate that under extremely high acceleration rates (e.g., 8x and 10x), the reconstruction accuracy of EGLAVarNet significantly outperforms existing state-of-the-art methods. Compared to the baseline model, the proposed network not only visually improves the structural fidelity of myocardial boundaries but also reduces parameter count and computational overhead by approximately 50%. It achieves an optimal balance between reconstruction accuracy and model complexity, providing a highly robust and low-deployment-cost new solution for the rapid clinical imaging of dynamic cardiac sequences.

KEYWORDS

Cardiac MRI, T1 Mapping, Deep Unrolling, Gated Linear Attention

CITE THIS PAPER

Chenqi Liu. Accelerated Cardiac MRI T1 Mapping Reconstruction Using Gated Linear Attention Network. Journal of Artificial Intelligence Practice (2026). Vol. 9, No. 1, 85-98. DOI: http://dx.doi.org/10.23977/jaip.2026.090111.

REFERENCES

[1] Look, D.C. and Locker, D.R. (1970) Time Saving in Measurement of NMR and EPR Relaxation Times. Review of Scientific Instruments, 41(2), 250–251. doi:10.1063/1.1684482.
[2] Deichmann, R. and Haase, A. (1992) Quantification of T1 Values by SNAPSHOT-FLASH NMR Imaging. Journal of Magnetic Resonance (1969), 96(3), 608–612. doi:10.1016/0022-2364(92)90347-A.
[3] Messroghli, D.R., Radjenovic, A., Kozerke, S., Higgins, D.M., Sivananthan, M.U. and Ridgway, J.P. (2004) Modified Look-Locker Inversion Recovery (MOLLI) for High-Resolution T1 Mapping of the Heart. Magnetic Resonance in Medicine, 52(1), 141–146. doi:10.1002/mrm.20110.
[4] Pruessmann, K.P., Weiger, M., Börnert, P. and Boesiger, P. (2001) Advances in Sensitivity Encoding with Arbitrary k-Space Trajectories. Magnetic Resonance in Medicine, 46(4), 638–651. doi:10.1002/mrm.1241.
[5] Griswold, M.A., Jakob, P.M., Heidemann, R.M., Nittka, M., Jellus, V., Wang, J., Kiefer, B. and Haase, A. (2002) Generalized Autocalibrating Partially Parallel Acquisitions (GRAPPA). Magnetic Resonance in Medicine, 47(6), 1202–1210. doi:10.1002/mrm.10171.
[6] Lustig, M., Donoho, D. and Pauly, J.M. (2007) Sparse MRI: The Application of Compressed Sensing for Rapid MR Imaging. Magnetic Resonance in Medicine, 58(6), 1182–1195. doi:10.1002/mrm.21391.
[7] Zhu, B., Liu, J.Z., Cauley, S.F., Rosen, B.R. and Rosen, M.S. (2018) Image Reconstruction by Domain-Transform Manifold Learning. Nature, 555(7697), 487–492. doi:10.1038/nature25988.
[8] Hammernik, K., Klatzer, T., Kobler, E., Recht, M.P., Sodickson, D.K., Pock, T. and Knoll, F. (2018) Learning a Variational Network for Reconstruction of Accelerated MRI Data. Magnetic Resonance in Medicine, 79(6), 3055–3071. doi:10.1002/mrm.26977.
[9] Shlezinger, N., Segarra, S., Zhang, Y., Avrahami, D., Davidov, Z., Routtenberg, T. and Eldar, Y.C. (2025) Deep Unfolding: Recent Developments, Theory, and Design Guidelines. arXiv preprint arXiv:2512.03768 (2025). doi:10.48550/arXiv.2512.03768.
[10] Sriram, A., Zbontar, J., Murrell, T., Defazio, A., Zitnick, C.L., Yakubova, N., Knoll, F. and Johnson, P. (2020) End-to-End Variational Networks for Accelerated MRI Reconstruction, in: Martel, A.L., Abolmaesumi, P., Stoyanov, D., Mateus, D., Zuluaga, M.A., Zhou, S.K., Racoceanu, D. and Joskowicz, L. (Eds.), Medical Image Computing and Computer Assisted Intervention–MICCAI 2020, Springer International Publishing, 2020, pp. 64–73. doi:10.1007/978-3-030-59713-9_7.
[11] Liang, Z.-P. (2007) SPATIOTEMPORAL IMAGINGWITH PARTIALLY SEPARABLE FUNCTIONS, in: 2007 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, 2007, pp. 988–991. doi:10.1109/ISBI.2007.357020.
[12] Yiasemis, G., Sonke, J.-J., Sánchez, C. and Teuwen, J. (2022) Recurrent Variational Network: A Deep Learning Inverse Problem Solver Applied to the Task of Accelerated MRI Reconstruction. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 732–741.
[13] Shen, Z., Zhang, M., Zhao, H., Yi, S. and Li, H. (2018) Efficient attention: Attention with linear complexities. arXiv preprint arXiv:1812.01243 (2018). doi:10.48550/arXiv.1812.01243.
[14] Zhao, Y., Zhang, Y. and Tao, Q. (2024) Relaxometry Guided Quantitative Cardiac Magnetic Resonance Image Reconstruction, in: Camara, O., Puyol-Antón, E., Sermesant, M., Suinesiaputra, A., Tao, Q., Wang, C. and Young, A. (Eds.), Statistical Atlases and Computational Models of the Heart. Regular and CMRxRecon Challenge Papers, Springer Nature Switzerland, 2024, pp. 349–358. doi:10.1007/978-3-031-52448-6_33.
[15] Adamson, P.M., Desai, A.D., Dominic, J., Varma, M., Bluethgen, C., Wood, J.P., Syed, A.B., Boutin, R.D., Stevens, K.J., Vasanawala, S., Pauly, J.M., Gunel, B. and Chaudhari, A.S. (2025) Using Deep Feature Distances for Evaluating the Perceptual Quality of MR Image Reconstructions. Magnetic Resonance in Medicine, 94(1), 317–330. doi:10.1002/mrm.30437.

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 26022
Visits: 746899

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.