Software development for quality assurance using ACR SPECT phantom

2012 
2530 Objectives Software development for cross-platform analysis of quality assurance (QA) using ACR SPECT phantom. Methods Manufacturing and placement of the ACR phantom was conducted per the ACR accreditation criteria. The acquisition and reconstruction parameters were properly controlled. Ultra Deluxe Jaszczak Phantom, 8 gamma cameras and 2 SPECT CT scanners were used. We implemented the image processing software using Interactive Data Language (IDL). This in-house software consisted of semi-automatic measures of the basic ACR phantom criteria as follows. First, we implemented resolution using cold rods. We summed several slices to construct a 2 cm thick image for visual inspection of detect ability of cold rods. Second, we implemented "contrast" using cold spheres. To this end, we implemented x & y profile plotting and measurement of minimum and mean values in that profile. Finally, we implemented uniformity. Since there is no gold standard of uniformity measurement of ACR phantom thus far, we used integral uniformity and coefficient of variation as used in routine planar gamma camera imaging practices. Results The software realized three fields of image resolution, contrast and uniformity that is required for ACR accreditation. In phantoms produced with Tc-99m, mean/SD of the integral uniformity was 21.3/8.1, and with Tl-201 was 22.9/4.5. The contrast values (i.e., ratio between mean/minimum values) were 5.58 and 7.34, respectively. The resolvable sizes of cold rods and spheres were 19.1mm and 10mm for Tc-99m, and were 9.3mm and 11.1mm for Tl-201. All images satisfied the ACR phantom criteria. We were able to assess the images without any complicated procedures. Moreover, our software exhibited compatibility on all available scanners in our center. Conclusions We successfully implemented QA software working in a semi-automatic manner and, in turn, worked more robustly than by manual methods. Moreover, our software exhibited compatibility on all available vendors9 scanners in our center. Research Support Asan Medical Cente
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []