Reconstruction strategies for gated myocardial perfusion SPECT: a phantom evaluation

The reconstruction strategy for gated myocardial perfusion SPECT can be stream-lined to reduce the computational demands of gated SPECT reconstruction, however, this may underestimate disease severity or increase false negative findings in the ungated qualitative image set. Methods: A cardiac phantom with defect insert was acquired using 360 degree SPECT and reconstructed using two alternative methods. The first, method A, produced data for qualitative assessment of perfusion by first ungating the raw data and then reconstructing. The second, method B, produced data for qualitative assessment of perfusion by reconstructing the raw gated file and then ungating the slices. Defect extent and severity were determined using CEqual software. Spatial measures of reconstructed luminal, diameter, myocardial diameter, wall thickness and defect size were also recorded. Results: The phantom results demonstrated that both the percentage extent of the defect and defect severity are reduced using method B compared to method A. Reconstruction using method B resulted in both an increase in total chamber diameter over method A and a reduction in the lumen diameter. Method B demonstrated a myocardial wall 20% thicker than method A. Method A was shown to underestimate the defect size by 5% while method B underestimates defect size by 30%. Conclusion: Minimising the computational demands of gated SPECT processing using method B may lead to false negative results. This may be particularly troublesome for non transmural defects. An evaluation of the clinical impact of this error is warranted.