ISMRM25 - AI-Based Acquisition & Reconstruction: Part II

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

AI-Based Acquisition & Reconstruction: Part II

Digital Poster

AI & Machine Learning

Tuesday, 13 May 2025

Exhibition Hall

14:30 - 15:30

Session Number: D-23

No CME/CE Credit

		Computer Number: 33 2763. Representation learning for ultra-low-field brain MRI super-resolution View Presentation Video X. Li, V. Lau, C. Man, A. Leong, Y. Zhao, E. Wu The University of Hong Kong, Hong Kong, China Impact: Enhancing image resolution and fidelity for ULF brain imaging at 0.055T using data-driven 3D deep learning approach. Potentially enable portable and point-of-care diagnosis.
		Computer Number: 34 2764. Fast hepatobiliary phase gadoxetate-enhanced imaging under breath-holding utilizing DL reconstruction (Sonic DL): preliminary experience View Presentation Video K. Sato, S. Tanaka, R. Murayama, Y. Takayama, A. Nozaki, X. Zhu, T. Cashen, A. Guidon, T. Wakayama, K. Yoshimitsu Fukuoka University, Fukuoka prefecture, Japan Impact: DLS-LAVA enables radiologists to obtain high-quality HBP images with reduced scan time, enhancing patient comfort and diagnostic precision compared to conventional LAVA. This advancement is especially valuable for patients with limited breath-holding capacity.
		Computer Number: 35 2765. Phase Inconsistency Problem of VarNet Reconstruction with External Calibration Data View Presentation Video K. Yang, W. Zhou, L. Mei, Y. Liu, M. Lyu Shenzhen Technology University, Shenzhen, China Impact: The proposed GRAPPA-based workaround solution to mitigate phase inconsistency improves the ability of E2E VarNet to utilize external calibration data, enhancing reconstruction quality and expanding its potential applications in research.
		Computer Number: 36 2766. Accelerated FLAIR imaging at 0.6T using T2W-guided multi-contrast deep learning-based reconstruction using a Zero-Shot approach View Presentation Video N. Jabarimani, C. Rao, E. Ercan, Y. Dong, N. Pezzotti, M. Doneva, M. van Osch, M. Staring, M. Nagtegaal Leiden University Medical Center, Leiden, Netherlands Impact: We showed that FLAIR and T₂-weighted scans can be used for contrast/style-based reconstruction methods, even when trained at 3T data directly applied to 0.6T data. Resulting improved image quality improves the usability of mid-field MRI.
		Computer Number: 37 2767. A probabilistic denoising diffusion-based framework for even higher accelerated quantitative MRI View Presentation Video P. Mayo, C. Pirkl, A. Achim, B. Menze, M. Golbabaee University of Bristol, Bristol, United Kingdom Impact: Our proposed approach enables the efficient use of Improved Denoising Diffusion Probabilistic Models for reconstructing highly accelerated quantitative MRI acquisitions, such as Magnetic Resonance Fingerprinting, leading to more accurate tissue parameter estimations.
		Computer Number: 38 2768. A Unified Reconstruction Framework for Arbitrarily Accelerated MR Imaging View Presentation Video Z. Jiang, K. Sun, D. Shen School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China Impact: We adopt feature representation transfer in the field of MRI reconstruction to address a practical issue largely overlooked by existing studies. Our adaptive reconstruction model can significantly simplify the deployment of MR reconstruction model and reduce the development costs.
		Computer Number: 39 2769. MRI partial Fourier deep learning reconstruction with implicit phase constraint View Presentation Video L. Zhang, S. Wang, H. Kutsuna, K. Shinoda, C. Wang Canon Medical Systems (China), Beijing, China Impact: Our method integrates implicit phase constraint into partial Fourier deep learning reconstruction network, it has been proved that the method works well in multi-anatomy, and it is expected the applications of partial Fourier with deep learning reconstruction can be expanded.
		Computer Number: 40 2770. Motion-Guided Deep Image Prior for 3D Real-Time Cine (M-DIP-3D) View Presentation Video C. Chen, M. Vornehm, M. Sultan, S. Arshad, Y. Han, R. Ahmad The Ohio State University, Columbus, United States Impact: The proposed method, M-DIP-3D, facilitates 3D real-time cine imaging from highly undersampled data. By directly learning the underlying motion and content variation manifold, M-DIP-3D produces images with minimal motion blur and real-time dynamics.
		Computer Number: 41 2771. The Role of Optimizing Loss Functions in Transfer Learning to Address Data Scarcity View Presentation Video S. Graf, W. Wohlgemuth, A. Deistung University Hospital Halle (Saale), Halle (Saale), Germany Impact: This study demonstrated the potential of optimizing transfer learning to adapt pre-trained models, even from different training settings, to new target-specific data, highlighting the great potential of cross-domain knowledge transfer and fine-tuning in addressing data scarcity.
		Computer Number: 42 2772. Stepwise hard constraint-biased Physics-informed Neural Networks for accurate Magnetic Resonance Electrical Properties Tomography (MREPT) View Presentation Video R. Qin, J. Yang, Z. Zhou, J. Gomez-Tames, S. Huang, W. Yu Department of Medical Engineering, Chiba University, Chiba, Japan Impact: This stepwise constraint-biased PINN approach could enable accurate MREPT without any ground truth, thus represent a step forward for clinical application of MREPT.
		Computer Number: 43 2773. RGB2qMRI: Can Deep Learning Models for Quantitative MRI be Trained with RGB Pictures? View Presentation Video S. Wang, A. Samadifardheris, J. A. Hernandez-Tamames, C. Pirkl, M. Vogel, L. Nuñez-Gonzalez, D. H. Poot, F. Wiesinger Erasmus MC, Rotterdam, Netherlands Impact: We demonstrated the feasibility of using widely available, non-anatomical RGB pictures to train deep learning models for qMRI, addressing the issue of limited training data, reducing costs of building training datasets, and avoiding biases toward specific health or disease types.
		Computer Number: 44 2774. K-space replacement layer for improving image sharpness in machine learning reconstruction View Presentation Video S. Z. K. Sajib, S. Bhave, S. Sharma Canon Medical Research, Cleveland, United States Impact: The proposed method has the potential to increase clinical throughput by reducing scan time while maintaining high image quality.
		Computer Number: 45 2775. Accelerating High Resolution 3D EPI with Deep Learning Reconstruction View Presentation Video N. Abad, S. Ahn, R. Brada, T. Sprenger, B. Fernandez, S. Banerjee, T. B. D. Yeo, T. Foo GE HealthCare, Technology & Innovation Center, Niskayuna, United States Impact: Deep learning based sparse image reconstruction can accelerate ultra-high resolution 3D EPI scans for brain imaging with acceleration factors ranging from 3-10, enabling disambiguation of clinically relevant fine features in various neuro imaging applications such as SWI, DWI and fMRI.
		Computer Number: 46 2776. Improved MRA Synthesis with KAN-based Network with Global MIP Guidance View Presentation Video D. Wang, S. Pasumarthi, A. Shankaranarayanan, G. Zaharchuk Subtle Medical Inc. , Columbia, United States Impact: The proposed method enables TOF-MRA synthesis from faster T1- and T2-weighted MRI images. The synthesized images closely match standard-of-care (SOC) TOF-MRA, accurately depicting vessels in major central areas, thus eliminating need for lengthy acquisition process of SOC-TOF-MRA in clinical settings.
		Computer Number: 47 2777. ZSSPI : Zero-Shot Scan Specific Parallel Imaging MRI Reconstruction using Attention based mechanism View Presentation Video J. Joo, H. Kim, T. Eo, H. Won, D. Hwang Yonsei University, Seoul, Korea, Republic of Impact: Our AKSM enables robust zero-shot MRI reconstruction by effectively utilizing undersampled k-space data. This breakthrough allows for significant scan time reduction without compromising image quality, potentially transforming clinical practice and making advanced MRI more accessible to patients.

Similar Session(s)