Design, development and performance evaluation of CT imaging in Inliview 3000 animal PET/SPECT/CT system

This work aimed to develop and evaluate the CT imaging performance in Nucmed Inliview 3000 animal PET/SPECT/CT system. Accurate geometrical calibration was implemented to enable the high resolution animal CT imaging. With the three ball bearings with known relative position, the projection data were acquired in helical scanning with 1333 views, 0.60 degree per view. The centroids of the bearings projection were fitted to the CT geometry model to estimate the geometrical parameters. FDK algorithm with Hilbert filter was used for 3-D reconstruction. The Graphics Processing Unit (GPU)-based acceleration was implemented on all the data pre-processing and FDK reconstruction. CT performances, including point spread function (PSF), modulation transfer function (MTF) and contrast discrimination function (CDF), were quantified with a uniform cylindrical organic glass phantom scanned in a helical orbit. Mouse experiments with helical scanning were also presented. The magnification ratio measured with calibration experiments was 1.8 at the center of FOV. With the calibrated parameters, the measured CT spatial resolution was 7.8 lp/mm @ 10% MTF. The contrast discrimination function was 2.06% with 1.5 mm size cubic box object. The mouse experiment showed this CT was feasible for the high resolution animal anatomic imaging. With GPU-base acceleration, the calculation time was about 10 minutes for a complete reconstruction of a 500 x 512 x 512 image from 4000 views of 1100 x 1700 projections. In conclusion, the animal CT in Nucmed Inliview 3000 system is adequate for practical animal imaging with high resolution. In future work, complete system performance evaluation will be performed on this system.