ISMRM25 - AI-Based Real-Time Imaging & Motion-Robust Strategies

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

AI-Based Real-Time Imaging & Motion-Robust Strategies

Oral

Acquisition & Reconstruction

Monday, 12 May 2025

310 (Lili-u Theater)

13:45 - 15:45

Moderators: Rasim Boyacioglu & Ecrin Yagiz

Session Number: O-08

No CME/CE Credit

13:45		0116. Self-supervised motion-compensated reconstruction for cardiac Cine MRI PDF Poster View Presentation Video S. Xu, A. Ghoul, K. Hammernik, J. Kuebler, P. Krumm, A. Lingg, D. Rueckert, S. Gatidis, T. Kuestner Medical Image and Data Analysis (MIDAS.lab), Department of Diagnostic and Interventional Radiology, University of Tuebingen, Tübingen, Germany Impact: We propose a fully self-supervised framework that enables high-quality reconstruction of cardiac Cine MRI acquired in a single breath-hold. The adaptability of this framework opens new research avenues for leveraging undersampled data and extends to other dynamic MRI modalities.
13:57		0117. AIM-ZS: Attention-driven Image reconstruction and Motion estimation using Zero-Shot self-supervised learning for dynamic MRI View Presentation Video N. Fujita, S. Yokosawa, T. Shirai, Y. Terada University of Tsukuba, Tsukuba, Japan Impact: This pre-training-free approach enables the widespread adoption of motion-corrected dynamic MRI without requiring large training datasets. It potentially expands clinical applications requiring high temporal resolution imaging and motion analysis and improving accessibility across different imaging protocols.
14:09		0118. Deep Learning-Based Joint Motion Correction, Reconstruction, and Segmentation for Free-Breathing Cardiac T1 Mapping View Presentation Video H. Chen, Z. Chen, J. Gao, C. Hu School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China Impact: The proposed method is expected to promote the clinical usability and reliability of cardiac T1 mapping, especially for patients who cannot tolerate repeated breath-holding MRI scans.
14:21		0119. Zero-Shot Self-Supervised Motion Corrected Cardiac T1 Mapping View Presentation Video M. Guastini, F. Zimmermann, J. Schulz-Menger, C. Kolbitsch, A. Kofler Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany Impact: The proposed method represents a valid alternative to supervised deep learning quantitative reconstruction methods, by employing the advantages of deep learning techniques without the need of target or training data, which is often unavailable.
14:33		0120. Motion-Guided Deep Image Prior for Dynamic Cardiac MRI View Presentation Video M. Vornehm, C. Chen, M. A. Sultan, S. M. Arshad, F. Knoll, R. Ahmad Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany Impact: Our method enables real-time free-breathing cine and free-breathing LGE imaging with high resolution and motion fidelity. It requires no training data and can be extended to other types of dynamic MRI acquisitions.
14:45		0121. Generative MR Multitasking: Bridging real-time and gated imaging with complex-valued cardiac encoding and neural subspace representations View Presentation Video X. Fang, A. Christodoulou David Geffen School of Medicine at UCLA, Los Angeles, United States Impact: This framework bridges traditional gating and real-time imaging approaches, and provides a flexible, scalable framework for multidimensional imaging. The complex-harmonic cardiac representation preserves both cardiac timing (phase) and beat-to-beat variability (amplitude), which may especially have future impact for arrhythmia patients.
14:57		0122. dDiMo: Domain-conditioned and Temporal-guided Diffusion Modeling for Accelerated Dynamic MRI View Presentation Video L. Zhang, I. Zhou, F. Liu Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, United States Impact: This work demonstrates the feasibility of a novel generative AI method for rapid dynamic MRI by leveraging temporal information and self-consistent k-t priors. Beyond its immediate applications, the method shows potential for generalization, adapting to inverse problems across various domains.
15:09		0123. Calibration-free DCE-MRI with Sub-second Temporal Resolution Using Interpretable Implicit Neural Representation PDF Poster View Presentation Video J. Feng, J. Chen, Y. Zhang, L. Feng, D. Liang, H. Wei Shanghai Jiao Tong University, Shanghai, China Impact: The proposed method paves the way for artifact-free, high-temporal-resolution DCE-MRI for clinical applications. Moreover, the proposed INR framework makes calibrationless MRI reconstruction feasible and interpretable, offering a powerful tool for a wide range of imaging challenges.
15:21		0124. End-to-end deep learning auto-navigation and reconstruction for accelerated free-breathing motion-resolved MRI View Presentation Video V. Murray, Y. Wen, G. Behr, O. Akin, A. Guidon, R. Otazo Memorial Sloan Kettering Cancer Center, New York City, United States Impact: The combination of deep learning auto-navigation and motion-resolved reconstruction enables fast and robust free-breathing abdominal MRI, which has the potential to reduce the number of repeat scans and increase efficiency compared to current clinical standards.
15:33		0125. Model-Agnostic Deep Learning Approaches for Dynamic Magnetic Resonance Thermometry Reconstruction View Presentation Video S. Zong, Y. Zhao, Q. Yang, H. Wang, G. Shen Fudan University, Shanghai, China Impact: The study's results enable faster and more accurate MR thermometry, allowing clinicians to better monitor and control temperature in thermal therapies. This advancement encourages further investigation into deep learning's role in MR applications, potentially extending to other phase-sensitive imaging techniques.

Similar Session(s)