Cape Town - 2026 ISMRM-ISMRT Annual Meeting and Exhibition - Diagnostic Utility of Advanced Techniques in MSK Applications

Diagnostic Utility of Advanced Techniques in MSK Applications

Digital Poster

Musculoskeletal

Wednesday, 13 May 2026

Digital Posters Row F

16:55 - 17:50

Session Number: 565-06

No CME/CE Credit

Similar Sessions

This session focuses on diagnostic evaluation of accelerated acquisition and Deep-Learning reconstructions methods for improved MSK MRI.

		565-06-001. Clinical Evaluation of Deep Learning Accelerated Lumbar T2-Weighted and Fat-Suppressed MRI Sequences Ran Lv, Nan Chen, Yijiang Huang, Hongtao Hou, Marcel Dominik Nickel, Yunzhu Wu, Guoqun Mao, Fuquan Wei Impact: This study enables 41.6% faster lumbar MRI with diagnostic-equivalent quality using DL-accelerated sequences, directly improving patient tolerance and workflow efficiency. It provides clinicians with a quantitative tool for early disc degeneration detection and facilitates clinical adoption of AI-reconstructed protocols
		565-06-002. Evaluation of a Deep Learning MR Reconstruction Model for Accelerating Clinical Knee Exams Emma Bahroos, Quin Lu, Michael Carl, Ajeetkumar Gaddipati, Alexandra Gersing, maggie fung, Thomas Link, Sharmila Majumdar University of California San Francisco, San Francisco, United States of America Impact: The findings of this study will inform the clinical viability of integrating deep learning-based reconstruction models into routine MRI knee exams which has the potential in improving workflow efficiency and patient experience.
		565-06-003. MRI in Clinical Practice: Diagnosis of Hepatic Invasion in a Large Abdominal Wall Sarcoma Using Dynamic-MRI Asako Yamamoto, Kenji Sato, Jungo Imanishi, Yoshinao Kikuchi, Hiroshi Imai, Chiaki Sato, Minami Hirasawa, Hiroshi Oba Teikyo University School of Medicine, Tokyo, Japan Impact: Dynamic-MRI with recent technological advancements enabled confident exclusion of hepatic invasion by a large abdominal wall sarcoma, leading to a downsized surgical approach. This technique offers a non-invasive, contrast-free alternative for dynamic tumor evaluation, with potential applicability across musculoskeletal oncology.
		565-06-004. MRI in Clinical Practice: Diagnosis of Osteosarcoma and Ewing’s Sarcoma in Diaphysis Zhendong Luo, Jiajing Huang, Jinjie Li, Qiyang Jiang, Zhiqiang Liu, Weiwei Yao, Jianxun Lv, Xinping Shen The University of Hong Kong - Shenzhen Hospital, Shenzhen, China Impact: Cortical notch sign improves OS/ES differentiation, guides treatment, drives imaging marker research, and optimizes patient outcomes.
		565-06-005. Clinical value of 3T MR nerve-bone fusion imaging in grading cervical neural foraminal stenosis using 3D-T2-FFE and FRACTURE Dongmei Jiang, Dejun She, Dairong Cao, Xiance Zhao The First Affiliated Hospital of Fujian Medical University, Fuzhou, China Impact: The use of 3T MR nerve-bone fusion imaging in clinical practice may facilitate a one-stop-shop, radiation-free, and more precise approach to comprehensively evaluate cervical neural foraminal stenosis.
		565-06-006. [18F]FDG PET-MRI for diagnostically challenging pain in patients with metal hip prostheses: a randomized controlled trial Jacob Mostert, Edwin Oei, Michael Ananta, Galied S Muradin, Pieter Bos, Biswal Sandip, Rianne van der Heijden Erasmus University Medical Center, Rotterdam, Netherlands Impact: [18F]FDG PET-MRI can identify pain sources in patients with chronic pain after total hip replacement leading to more targeted treatment. Preliminary results show no improved clinical outcome, but more research is needed.
		565-06-007. Lumbar Spine Imaging at 5.0 Tesla with Deep Learning-based Reconstruction: Improvement of Efficiency and Image Quality Lixin Du, Pan Wang, Jing Yang, Hai Lin Shenzhen Longhua District Central Hospital, Shenzhen, China Impact: The results demonstrate that DLR can significantly reduce MRI scan times while improving image quality, enhancing patient comfort and clinical workflow. This study enables further exploration into AI-driven imaging techniques and their broader implications for patient care and diagnostic accuracy.
		565-06-008. Deep Learning-based Reconstruction Enhances Image Quality and Diagnostic Performance in 5.0 Tesla Knee MRI Lixin Du, Pan Wang, Jing Yang, Hai Lin Shenzhen Longhua District Central Hospital, Shenzhen, China Impact: This study establishes the clinical value of DLR in 5.0T knee MRI, confirming its significant SNR improvement and arthroscopy-validated diagnostic superiority, thereby supporting the expanded application and workflow optimization of ultra-high field MRI.
		565-06-009. Quantitative DCE-MRI–Guided Dose Escalation for Soft Tissue Sarcoma: A Feasibility Study Comparing IMRT, IMPT, and SPArc  xiaoda cong, zerun zhang, Wei Huang, xuanfeng ding Beaumont Hospital, Royal Oak, United States of America Impact: This feasibility study shows that quantitative functional imaging-guided SPArc radiotherapy planning can achieve an optimal balance between dose escalation and normal-tissue sparing and has the potential to be integrated in clinical practice to facilitate personalized radiotherapy.
		565-06-010. Deep learning-accelerated T2-weighted SPACE for cervical spine MRI: faster acquisition with preserved image quality Thierno Diallo, Kai Falko Kästingschäfer, Dominik Paul, Marcel Dominik Nickel, Ralph Strecker, Matthias Jung, Fabian Bamberg, Jakob Weiß, Alexander Rau, Maximilian Frederik Russe University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany Impact: Deep learning reconstruction enabling 47% faster cervical spine T2w SPACE acquisitions with preserved diagnostic confidence could reduce discomfort from prolonged positioning, improve radiology throughput, and facilitate high-resolution imaging in claustrophobic patients or those with difficulty remaining still during longer scans.
		565-06-011. Can contrast-enhanced images be synthesized from non-contrast quantitative MRI in knee MRI? Olli Nykänen, Marko Nikki, Victor Casula, Riccardo Lattanzi, Mikko Nissi, Miika T Nieminen, Mika Nevalainen University of Oulu, Oulu, Finland Impact: The present study shows that by deep learning, natural-looking contrast-enhanced MRI images can be synthesized from non-enhanced MRF data. However, as the images were not clinically viable, the study underlines that deep learning-synthesized images should be used thoughtfully.
		565-06-012. Development and Clinical Validation of a Whole-Ankle MRI Scoring System for Osteoarthritis: Reliable Phenotyping and Prognost Tao Li, QINGMEI Yang, WeiHong Huang, DeXuan Chen, YingHua Zhao The Third Affiliated Hospital Southern Medical University, Guangzhou, China Impact: This whole-ankle MRI score is a standardized imaging biomarker for ankle osteoarthritis. It distinguishes structural vs inflammatory phenotypes and predicts long-term pain trajectory, enabling individualized management and targeted clinical trial enrichment.
		565-06-013. Advantages of 2.5D FS-PDWI for Knee MRI at 5T: A Comparative Study with Conventional 2D Imaging Yunyun He, Liang Zhou, Hong Wang, Hui Huang, Wenze Wu, Sibin Liu Jingzhou Hospital Affiliated to Yangtze University, Jingzhou, China Impact: The 2.5D technique at 5T provides a clinically practical solution that acquires high-resolution, sharp knee images with thin-slice capability while avoiding the typical SNR penalty, thereby enhancing the detection and characterization of subtle cartilage lesions and other fine anatomical details.