Four-Dimensional Magnetic Resonance Fingerprinting (4D-MRF) Using Motion Compensation and Alternating Direction Method of Multipliers (ADMM) for Liver Cancer Radiotherapy
C Liu1, T Li1, P Cao2, E Hui1, Y Wong1, Z Wang2, H Xiao1, W Li1, X Teng1, J Zhang1*, A Cheung2, H Lee2, M Ying1, J Cai1, (1) The Hong Kong Polytechnic University, Hong Kong, China (2) The University Of Hong Kong, Hong Kong, China
Presentations
TU-D930-IePD-F5-4 (Tuesday, 7/12/2022) 9:30 AM - 10:00 AM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 5
Purpose: Magnetic resonance fingerprinting (MRF) has been developed for simultaneous mapping of tissue properties (T1, T2, proton density) in a single MRI acquisition. This study aims to develop a novel 4D-MRF technique for abdominal multi-parametric quantitative imaging and to evaluate its clinical efficacy in liver cancer radiotherapy.
Methods: The 4D-MRF reconstruction was formulated as a motion-compensated low-rank inverse problem, which was solved by the alternating direction method of multipliers (ADMM). The free-breathing MRF signals were retrospectively sorted into five respiratory phases based on the amplitude of the breathing curve. The inter-phase motion was calculated by a free-form deformation algorithm. The ADMM incorporated the deformation field and removed the motion artifact iteratively. The 4D-MRF technique was validated on digital motion phantom, three health volunteers, and nine liver cancer patients. The 4D-MRF data was acquired by a 3T system (SIGNA Premier, GE healthcare) using a customized multi-slice FISP sequence with 1000 dynamics. Patients’ clinical scans, including planning CT and contrast-enhanced MRI (CE-MRI), were collected in a prospective clinical trial.
Results: The 4D-MRF technique successfully reconstructed quantitative 4D tissue property maps with accurate organ boundaries. The motion-compensated ADMM reconstruction showed improved mapping accuracy by 24.12%/20.81% for T1/T2 in the phantom study. In volunteer and patient study, the quantitative maps agreed with reference values and the difference was within 10%/20% for T1/T2 values for tissues of interest. The T1 map had an improved contrast-to-noise ratio (CNR) compared with planning CT and had a comparable CNR with CE-MRI, which were 12.660 ± 5.794, 3.743 ± 0.988, and 14.653 ± 0.125, respectively.
Conclusion: This is the first patient study on 4D-MRF for liver cancer radiotherapy. 4D-MRF demonstrated high accuracy in both tissue property quantitation and motion measurement. It holds great potential to facilitate radiotherapy-related clinical applications, including liver tumor delineation, treatment planning, and motion management.
Funding Support, Disclosures, and Conflict of Interest: This research was supported by research grant of General Research Fund (GRF 15102219), the University Grants Committee, the Hong Kong Special Administrative Region, China.
Keywords
Quantitative Imaging, Image-guided Therapy, Patient Movement
Taxonomy
IM/TH- MRI in Radiation Therapy: Development (new technology and techniques)
Contact Email