Differential effect of 180 degrees and 360 degrees acquisition orbits on the accuracy of SPECT imaging: quantitative evaluation in phantoms

Abstract

A circular 180 degrees acquisition orbit is considered standard for cardiac SPECT imaging. Theoretically, a 360 degrees acquisition orbit is preferred because of more complete Fourier spectral information on projection data. The differential effect of 180 degrees and 360 degrees acquisition orbits on image accuracy (homogeneity and defect size) was assessed quantitatively in phantom studies.

Methods: SPECT imaging with a dual-head gamma camera was performed on normal cardiac phantoms filled with a (99m)Tc solution, using 180 degrees and 360 degrees circular acquisition orbits. The phantoms were placed in the center of the orbit and at 5, 10, and 15 cm off center. Fillable defect inserts of different sizes were placed in the phantom to simulate myocardial perfusion defects. The homogeneity of count distribution in short-axis slices of the normal phantom was analyzed as the percentage of variability. Defects were quantified as a percentage of the entire phantom volume using circumferential count profiles and normal reference profiles.

Results: When normal phantoms were placed in the center of the orbit, percentage variability was not different whether a 180 degrees or 360 degrees acquisition orbit was used (4.2% +/- 0.1% vs. 4.4% +/- 0.2%, P = not statistically significant). However when normal phantoms were placed off center, SPECT imaging with a 180 degrees acquisition orbit showed increasing inhomogeneity, both visually and quantitatively (e.g., percentage variability for the 15-cm off-center position was 10.8% +/- 0.1% (P < 0.0001). SPECT imaging with a 360 degrees acquisition orbit showed similar homogeneity visually and quantitatively, whether the phantom was placed in or off the center (e.g., percentage variability for the 15-cm off-center position was 4.6% +/- 0.5%, P = not statistically significant). Quantification of phantom defects acquired with a 180 degrees orbit showed increasing overestimation of defect sizes with increasingly off-center positions. Quantification of phantom defects acquired with a 360 degrees orbit showed no effect from progressively off-center positions, although phantom defect sizes were mildly underestimated.

Conclusion: SPECT images acquired with a 180 degrees orbit may have significant erroneous inhomogeneity and overestimation of defect size, in particular when the target object is off the center of the orbit, as is commonly seen in clinical cardiac imaging. In contrast, SPECT images acquired with a 360 degrees orbit may provide more accurate quantitative information.