Measuring Spatiotemporal Resolution in Real-Time MRI

Motivation: Real-time MRI can provide powerful insights into dynamic processes, but practical experimental limitations have led to the widespread use of undersampled data. While advanced reconstruction methods can mitigate undersampling artifacts, these methods are unlikely to be perfect, and their rigorous validation has been a longstanding open problem. Goal(s): To introduce a new reference-free approach for evaluating real-time MRI results. Approach: We introduce a framework for measuring spatiotemporal resolution in real-time MRI, based on the propagation of spatiotemporal perturbations through image reconstruction. Results: The proposed approach is sensitive to spatiotemporal resolution features, and provides valuable new information for the interpretation of real-time MRI results. Impact: The proposed framework enables measurement of spatiotemporal resolution, providing new information that is important for the interpretation of real-time MRI results, and can also be useful for the development/tuning of acquisition and reconstruction methods.