ROI-based polar maps (33 ROIs) were employed to evaluate quantitatively stress/rest myocardial 201TI SPECT in 108 patients with angiographically proven coronary heart disease (CHD) in comparison with 30 controls. Sensitivity in detecting a CHD with stenoses of > 50% of luminal diameter was determined versus normal regional values (± 2.5 SD) employing vitality (VI) and wash-out corrected redistribution (RD). The method was evaluated referring to the severity of the disease, to the number of ROIs displaying changes [(a) 1 ROI, (b) >2 ROIs], to validity of VI, RD or a combination thereof, and for specificity. Wash-out values were found to depend on degree of stress individually achieved and thus were not used as a threshold criterion. Sensitivity in supply areas with old myocardial infarctions was 95% (a) and 86% (b), resp. With no infarction, it was 96% (a) and 79% (b), resp. VI in stenosis > 75% was more sensitive than RD. However, combined evaluation of VI and RD yielded sensitivities from 91-100% (a) and 77-94% (b), resp. for different main supply areas. In stenosis < 50% with normal VI, RD was positive in 18-31 %. Specificity turned out to be 91 % (a) and 97% (b), resp. We conclude that the method presented is reliable to quantify numerically 201TI kinetics in myocardial SPECT, aimed at detecting and describing CHD.
Quantitative Evaluation of Snapshot Graphs for the Analysis of Temporal Networks
