Analytical Reconstruction of Periodically Deformed Objects in Time-resolved CT
To address the low radiation utilization efficiency and phase-wise information fragmentation in time-resolved CT reconstruction of periodically moving organs (e.g., heart, lungs), this paper proposes two novel analytical joint reconstruction paradigms. For the first time within an analytical framework, our approach fully integrates projection data across the entire motion cycle, enabling cross-phase motion-coupled modeling and simultaneous dose–noise optimization. The method is grounded in motion-constrained analytical backprojection theory and periodic deformation-geometry modeling, balancing structural fidelity and noise suppression. Synchrotron-based micro-CT experiments demonstrate substantial random noise reduction while preserving edge sharpness; at equivalent image quality, radiation dose is reduced by approximately 40%. The implementation code is publicly available.