Optimizing Liver MRI Using Breath-Holding With and Without Oxygen

This is a within-subject, randomized crossover interventional study conducted at UCSF China Basin Imaging Center using a 3 Tesla Magnetic Resonance Imaging (3T MRI scanner). The study aims to optimize abdominal MRI protocols by evaluating the impact of breath-holding techniques, with and without oxygen supplementation, on motion artifacts and overall image quality. Each participant will undergo non-contrast abdominal MRI scans using two breath-hold conditions: (1) Functional Residual Capacity (FRC)/end-expiration without oxygen and (2) FRC/end-expiration with preoxygenation. The order of these conditions will be randomized to control for potential order effects, and each participant will serve as their own control to reduce inter-individual variability. T2-weighted and Magnetic Resonance Cholangiopancreatography (MRCP) sequences will be acquired during each breath-hold trial. MRI-compatible pulse oximetry will be used to monitor oxygen saturation and pulse rate throughout scanning. A radiology research nurse will be present if additional monitoring is needed based on the participant's pre-screening or intra-scan findings. Participants will receive training on breath-hold procedures prior to imaging to improve consistency and comfort during scanning. Two board-certified radiologists, blinded to the breath-hold condition, will independently assess each image set for motion artifacts and overall image quality using a standardized 5-point grading system. The grading will range from very poor (1) to very good (5) based on sharpness, visibility of anatomical details, and presence of blurring or signal loss. Inter-rater reliability will be evaluated using the Intraclass Correlation Coefficient (ICC). Statistical analyses will include repeated measures ANOVA to compare image quality scores across breath-holding conditions. Descriptive analyses will assess breath-hold durations and participant characteristics (e.g., age, gender) in relation to image quality and safety outcomes, including oxygen saturation and self-reported comfort. This study involves minimal risk, does not utilize contrast agents, and is intended to support the development of motion-reduced, efficient liver MRI protocols, particularly for living liver donor evaluation. Data collected during the study will not be used for clinical decision-making.