Studies On Approximate Reconstruction Algorithm For Multi-source Multislice Spiral CT

Spiral CT is an important diagnostic tool and is widely used in medical field. However, people still hope that CT could scan faster, collect more projection data and have more flexible scan locus. In this paper, we propose a new type of CT configuration, which consists of several X-ray sources. Each X-ray source is equipped with a multi-row detector. During the scan, all sets of sources and detectors rotate simultaneously and follow the helical scan locus. The pitch of helical locus can be fixed or variable depending on the requirements for applications. Different multi-source approximate reconstruction algorithms are supported for both scan modes. And in cardiac imaging, an ECG related multi-source 180°MCI (Multislice Cardio Interpolation) is given as well. We also define two digital phantoms, simulating the diffusion of contrast media in blood vessels and the beat of heart. The formulas to calculate the projections of phantoms are provided. The simulation results of the algorithms and the comparisons between the single-source spiral CT and the multi-source spiral CT show that our new CT model scans faster and has higher resolution in Z direction. It can also suppress the motion artifacts with the increment of the number of X-ray sources. And in the application of CTA (Computed Tomography Angiography), the multi-source spiral CT with variable pitch can chase contrast media more flexibly and more effectively.