Using attenuation-corrected images in myocardial perfusion scintigraphy reduces the need for a rest study, both for computerized and manual interpretations

Objectives Attenuation in myocardial perfusion scintigraphy (MPS) may require comparison of stress and rest images to distinguish perfusion abnormalities from artefacts. If a stress study is considered normal, there is no need for a rest study. The purposes of the study were to 1) compare the accuracy of computerized determination of the need for a rest study when using non-attenuation corrected (NC) images or attenuation-corrected (AC) images and 2) compare the evaluations of physicians determining the need for a rest study when using NC images or NC+AC images.

Methods 1261 patients admitted to 99Tc MPS were included. Stress studies were interpreted by three physicians who judged each study as “no rest study necessary” or “rest study necessary” (evaluating NC+AC images). Agreement between 2 out of 3 physicians was considered gold standard. The EXINI heartTM software package was used to interpret the AC and NC images separately for each patient. For the comparison between the manual interpretations, the evaluations from the physicians (NC only vs. NC+AC) were used. McNemar was used for statistical analysis.

Results For the computerized determination of the need for a rest study, the sensitivity for NC only was 78%, and 86% for AC images. Specificity was 83% and 80%, respectively and accuracy 80% and 84%, respectively (p<0.0001). When comparing the evaluations from the physicians (NC only and NC+AC, respectively), there were 236 discordant pairs, were the physicians wanted a rest study for NC+AC but not for NC in 11 cases and they wanted a rest study for NC but not for NC+AC in 225 cases (p<0.0001).

Conclusions AC images improves the detection of normal studies, thus the need for a rest study decreases. This approach could improve laboratory efficiency and reduce the radiation dose to the patient