Clinical implication of down-scatter in attenuation-corrected myocardial SPECT

Interpretation of myocardial perfusion single photon emission computed tomography (SPECT) studies is hampered by attenuation artifacts. Attenuation correction methods with simultaneous emission and transmission are now commercially available. However, it has been observed in clinical practice that attenuation correction without down-scatter correction in a 1-day rest/stress myocardial perfusion protocol may lead to serious interpretation errors. Therefore the aim of this study was to study errors resulting from down-scatter under realistic conditions, thus providing a background for the assessment of further corrections.

Forty-six patients underwent myocardial perfusion scintigraphy in a 1-day technetium 99m-tetrofosmin rest-stress SPECT protocol, with a moving 153Gd line-source device for attenuation correction without down-scatter correction. Short-axis slices were quantified as inferior/anterior, septal/lateral, and apical/remainder count ratios. The changes at rest (350 MBq) and exercise (900 MBq) induced by attenuation correction were studied. Attenuation correction gave differences in apparent perfusion between rest and exercise not seen before correction. The gender differences in inferior-anterior ratio were greatly reduced after correction at rest but remained at exercise. A torso phantom study indicated that these results were due to under-correction at exercise because of down-scatter.

Down-scatter results in an underestimation of attenuation in simultaneous emission and transmission, if not accurately accounted for. Particularly, a high-dose study compared with a low-dose study, as in the 1-day protocol, might cause serious interpretation errors.