Background. Photon attenuation is a major cause of artifacts on single-photon emission tomographic imaging. Methods and Results. To study a new method to perform photon attenuation correction (AC), we used a cardiac phantom filled with 99mTc and imaged it (1) without extrinsic attenuation or defects, (2) with extrinsic attenuation but without defects, (3) without extrinsic attenuation but with defects involving 10% of the myocardial volume and activity ranging from 0% to 75% of maximum, and (4) with both extrinsic attenuation and defects. Transmission and emission images acquired with a dual-head single-photon emission computed tomographic system with 153Gd line sources were processed by iterative maximum-likelihood-expectation-maximization and were evaluated both qualitatively and quantitatively. The small defects were readily identified both before and after AC. Mean count activity (percent of maximal activity) of the segments with overlying extrinsic attenuation but without defects was only 56% ± 4% without AC but increased to 86% ± 4% with AC (p < 0.0001). Without AC, the count activities in the defects with overlying extrinsic attenuation were lower than the actual defect activities, but AC resulted in better approximation of actual defect activities in all but the most severe (0% tracer activity) defects. Conclusion. This new AC method provided an improved estimation of actual myocardial count activity. Even small defects with mild reduction in tracer activity were still identifiable after AC.
- Attenuation correction
- Single-photon emission computed tomography
ASJC Scopus subject areas
- Cardiology and Cardiovascular Medicine