Develop and Optimize 5DCT Imaging Simulation and Reconstruction Methods.
Abstract
Purpose
To study the cardiac and respiratory heart motion together. Cardiorespiratory motion management is a significant topic in stereotactic arrhythmia radiotherapy (STAR). Cardiorespiratory motion consists of the heart beating and respiration. Commonly studied separately using respiratory 4DCTs and cardiac 4DCTs, these two types of motion are not truly independent, and it is crucial to study them together. For example, intrapulmonary pressure variations during the respiratory cycle significantly affect heart motion. However, there is currently no 5D (3D + respiratory + cardiac) imaging method to support such studies. To fill the gap, we developed a novel 5DCT imaging method and tested the feasibility using digital phantom simulation.
Methods
Helical 5DCT scans were simulated using 4D XCAT, following real CT scanner hardware and scanning parameters, producing instantaneous CT volumes, sinogram projections, and corresponding respiratory and cardiac signals at 3280 Hz. A novel 5DCT reconstruction algorithm was developed. The steps to reconstruct each 2D slice of a 3DCT volume for a specific cardiac and respiratory phase were 1) sinogram data filtering by cardiac phase, respiratory phase, and couch position, 2) helical-to-fanbeam projection data rebinning, 3) sinogram data linear-interpolation in the 2D space of cardiac and respiratory phases, and 4) 2D filtered backprojection. The sinogram interpolation step was designed to handle data scarcity and avoid axial slice misalignments in reconstructed 3DCTs.
Results
A 49-second helical scan of pitch = 0.02 was simulated to cover the whole heart longitudinally. The 5DCT reconstruction algorithm was successfully implemented to produce 5D images of higher quality than clinical 4DCTs, with slice misalignments eliminated by the sinogram interpolation method.
Conclusion
The 5DCT simulation and reconstruction pipeline demonstrated that 5DCT imaging was feasible. Further studies include the 5DCT imaging protocol optimization, reconstruction algorithm optimization, and physics phantom and patient imaging studies.