ISMRM25 - AI-Based Acquisition & Reconstruction: Part I

ISMRM & ISMRT Annual Meeting & Exhibition • 10-15 May 2025 • Honolulu, Hawai'i

AI-Based Acquisition & Reconstruction: Part I

Digital Poster

AI & Machine Learning

Tuesday, 13 May 2025

Exhibition Hall

13:30 - 14:30

Session Number: D-22

No CME/CE Credit

		Computer Number: 49 2626. DCSOlve-MR: Adaptive Integration of Generative AI and Compressed Sensing for High-Fidelity MRI Reconstruction View Presentation Video M. Huang, D. Schmidt, Y. Zhao, A. Petrovic, R. Bammer Monash University, Melbourne, Australia Impact: By leveraging the strengths of both generative AI and compressed sensing (CS), the proposed method faithfully restores images from undersampled measurements, achieves high-fidelity MRI reconstructions, and does not obfuscate the diagnostic quality of the scan or downstream AI tasks.
		Computer Number: 50 2627. Fast Cardiac T1 Mapping in Three Heartbeats with A Transformer Integrated Network View Presentation Video H. Li, Q. Liu, H. Shi, H. Gu, Y. Dong, C. Zhu, B. Li, Z. Zhong, G. Dan, Z. Chen, Y. Liu, Q. Liu, Y. Ye, J. Xu, J. Hou, F. Fang, Y. Zhu, S. Liu, H. Wang, D. Liang, Y. Zhou UIH America Inc, Houston, United States Impact: The approach achieves cardiac T1 mapping within three cardiac cycles, clinically reducing the breath-hold time and causes less discomfort to patients. Meanwhile, the integrated network takes advantages of current deep learning methods and significantly improves the reconstruction quality.
		Computer Number: 51 2628. Deep learning accelerated Sandwich for 3D multi-channel B1+ mapping at 7T in less than 10 seconds View Presentation Video N. Pato Montemayor, J. Philippe, J. Kent, A. Hess, A. Klauser, E. Sleight, L. Bacha, T. Di Noto, B. Maréchal, P. Liebig, J. Herrler, D. Nickel, R. Heidemann, J-P Tiran, T. Kober, T. Hilbert, T. Yu, G. F. Piredda Siemens Healthineers International AG, Lausanne, Switzerland Impact: This study demonstrates a deep learning-based method for rapid B₁⁺ mapping in ultra-high field MRI, significantly reducing acquisition time to under 10 seconds while maintaining accuracy. The approach enhances the efficiency of parallel transmission, facilitating clinical applications at 7T.
		Computer Number: 52 2629. Self-supervised motion-corrected reconstruction for single heartbeat cardiac cine MRI at 1.5T and 0.55T using PINNs and Deep Image Prior View Presentation Video T. Catalán, R. De la Sotta, R. Botnar, F. Sahli-Costabal, C. Prieto Millennium Institute for Intelligent Healthcare Engineering, Santiago, Chile Impact: The proposed approach enables TR-resolved single heartbeat cardiac cine at both 1.5T and 0.55T. This acceleration can be leveraged to reduce scan time, for example allowing more slides in the same breath-hold.
		Computer Number: 53 2630. 8-fold Accelerated TSE MRI: First Clinical Results of an End-to-End DL Reconstruction Method for Combined SMS and Parallel Imaging Acceleration View Presentation Video Y. Leonhardt, J. Vosshenrich, M. Mostapha, G. Koerzdoerfer, E. Raithel, M. Nadar, M. Bruno, J. Fritz Grossman School of Medicine, New York University, New York, United States Impact: First clinical results for 8-fold accelerated knee MRI using a dedicated end-to-end DL reconstruction approach for combined SMS and PI acceleration demonstrate image quality comparable to slower standard-of-care MRI, holding promise for significantly faster clinical TSE MRI in the future.
		Computer Number: 54 2631. Benchmarking 3D Multi-Coil NC-PDNet MRI Reconstruction View Presentation Video A. Tanabene, C. G.R, A. Massire, M. Nadar, P. Ciuciu Siemens Healthineers, Courbevoie, France Impact: Achieving fast, high-quality reconstructions with reasonable GPU memory usage demonstrates our approach's viability for clinical research. Benchmarking the GoLF-SPARKLING trajectory against established non-Cartesian baselines in 3D multi-coil settings validates its future application in more challenging experiments, notably at higher resolutions.
		Computer Number: 55 2632. Consistent MRI Reconstruction with Diffusion Models and Sequential Monte Carlo View Presentation Video W. Jiang, W. Song, Y. Gao, N. Ye, F. Liu, H. Sun The University of Queensland, Brisbane, Australia Impact: This method enhances MRI reconstruction consistency, increasing reliability for clinical use and establishing a foundation for broader adoption in diagnostic imaging and medical diagnostics.
		Computer Number: 56 2633. Motion-Corrected Deep-Learning Reconstruction Framework for 3D Whole-Heart MRA at 0.55T View Presentation Video M. Paredes, A. Phair, A. Fotaki, R. Botnar, C. Prieto Millennium Institute for Intelligent Healthcare Engineering, Santiago, Chile Impact: The proposed approach validates MoCo-MoDL feasibility at 0.55T, enabling 7-fold undersampled, non-rigid motion-corrected CMRA with 1.36min acquisition and 42s reconstruction, showing promise for clinical implementation in low-field environments and making MRI more attainable and cost-effective.
		Computer Number: 57 2634. Self-Supervised Pretraining of Joint Acquisition and Reconstruction for Fast Quantitative MRI H. Shang, S. Jia, D. Liang Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, shenzhen, China Impact: Large-scale pretrained end-to-end model can be an alternative to long standing Fourier Transform foundation for qMRI. The new capability of contemporary spatial-physical encoding opens up new possibility to push current speed limit of qMRI.
		Computer Number: 58 2635. Accelerating DT-CMR with Variable CAIPIRINHA Shifts and Artefact-Aware AI-based SMS reconstruction View Presentation Video M. Tänzer, E. j. Lim, H. (. Qiu, A. Scott, P. Ferreira, D. Rueckert, G. Yang, S. Nielles-Vallespin Imperial College London, London, United Kingdom Impact: By substantially reducing acquisition times while preserving diffusion map quality, this work helps bridge the gap towards clinical adoption of DT-CMR, advancing non-invasive assessment of cardiac microstructure for diagnosing and monitoring of cardiac disease.
		Computer Number: 59 2636. DCIPSR: A Self-supervised Networks with Signal Regularization for Multi-Contrast MRI Reconstruction View Presentation Video H. Wei, Z. Li, J. Zou, H. Meng, P. Ma, X. Liu, R. Li Tsinghua University, Beijing, China Impact: We proposed a self-supervised MCMRI reconstruction method with Densely Connected Image Prior(DCIP) and Signal Regulation(SR). The success of DCIPSR under high undersampled rate indicates the potential to reconstruct MCMRI when large training dataset is unavailable.
		Computer Number: 60 2637. Dep2SMS: Simultaneous Multi-Slice Reconstruction via Deep Learning with Auxiliary Depth Camera Guidance View Presentation Video M. Song, X. Hao, F. Qi, Z. Guo, Y. Li, B. Qiu University of Science and Technology of China, Hefei , China Impact: We novelty introduce the depth camera into the MRI system to capture contour to assist SMS reconstruction. We utilize the Mamba-based framework with intra-patch convolution and linear-complexity long-range attention for SMS reconstruction to capture fine structural and global texture features.
		Computer Number: 61 2638. Multi-Contrast MR Imaging Acceleration: Transform 2D Low-Resolution into High-Resolution 3D Images with Auxiliary Contrast in 3D Acquisition View Presentation Video Z. Zhang, Z. Zhou, L. Xiang, X. Song, X. Wei, Y. Li Subtle Medical Inc, Guildford, United Kingdom Impact: The developed approach allows multi-contrast brain low-resolution 2D scans with an auxiliary high-resolution 3D reference scan to produce multi-contrast high-resolution 3D images. Clinical evaluation on synthesized 3D brain images and extension to other applications may be worth further investigation.
		Computer Number: 62 2639. Denoising Diffusion Probabilistic Model with Dual-domain Entropy and Stochastic Differential Equations on High Undersampling Reconstruction View Presentation Video Z. Li, Y. Yang, J. An, C. Ling, Z. Wang, Y. Zhuo, R. Xue, Z. Zhang State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China Impact: The proposed EDU-DDPM significantly improves MRI reconstruction at high subsampling factors, outperforming DDPM on fastMRI and compressive sensing on 7T TOF-MRA data. This advancement enhances fidelity of reconstruction.
		Computer Number: 63 2640. Low-field MRI reconstruction with hourglass diffusion model and posterior sampling strategy View Presentation Video Y. Lian, J. Zhang, H. Guo Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Biomedical Engineering, Tsinghua University, Beijing, China Impact: An efficiency deep learning based method to accelerate the acquisition in low-field MRI and improve the image SNR
		Computer Number: 64 2641. Accelerated Quantitative Magnetization Transfer Mapping with Deep Learning (MTAcqNet) View Presentation Video Y. Wang, J. Lo, J. Athertya, Y. Ma Department of Electrical and Computer Engineering, University of California, San Diego, La Jolla, United States Impact: The proposed MTAcqNet accelerates MT data acquisition and enables faster macromolecular proton fraction mapping using quantitative MT modeling. By reducing scan time, it facilitates rapid clinical translation of quantitative MT imaging, providing valuable insights into macromolecular content in the brain.

Similar Session(s)