Reproducibility of ring artifacts visualized in Jaszczak phantom imaging.

Objectives: Nuclear Medicine imaging is reliant on the accuracy of the instrumentation used for examinations. Quality control is an established set of corrections and analyses that validate camera performance. Jaszczak phantom imaging provides performance information on SPECT systems including spatial resolution, contrast limits of detection, and reconstructed uniformity. Routinely, the phantom is only imaged once and the reconstructed image set visualized to assess these parameters. Given the inherent noise of SPECT imaging, it is possible that a ring artifact visualized on one scan may not be seen on a repeated scan, due to different counting statistics of the two acquisitions. However, if the ring artifact was stemming from poor system uniformity, it would be reproducible on multiple scans in addition to being visible in differing locations in the phantom itself if the phantom were repositioned. The objective of this study was to evaluate if a visible ring artifact in Jaszczak imaging was originating from poor camera performance or the non-optimal statistical nature of SPECT imaging.

Methods: A Jaszczak phantom was filled with approximately 30 mCi of uniformly distributed Tc99m, and imaged five times consecutively using the ACR standard protocol for SPECT systems. The phantom was then shifted 20 mm in the axial direction and imaged five more consecutive times. These scans were reconstructed and visually assessed.

Results: Although the artifact appeared with slight variations on each scan, reconstructed Jaszczak images demonstrated a reproducible ring artifact near the center of the phantom on the initial phantom scans. After shifting the phantom, the second run of reconstructions demonstrated a ring artifact in each scan (again appearing with slight variations), but in a different cross-sectional slice of the phantom, located approximately 6 slices inferiorly to the original ring artifact position (20mm shift, 3.3mm/slice = approximately 6 slices).

Conclusion: Ring artifacts visualized in SPECT imaging can be caused by non-uniform system sensitivity and non-optimal intrinsic/extrinsic flood corrections. Due to the fact that the ring artifact was reproducible after multiple imaging runs before and after moving the phantom, it was determined that the ring artifact stems from poor system uniformity, rather than statistical variations. Further investigation involving new corrections and repeat Jaszczak imaging is recommended.