Thallium myocardial perfusion scintigraphy: influence of perfusion, scatter, and photon energy on the detection of lesions.

Experimental studies have been made of lesion detectability in myocardial perfusion studies using thallium-201. A series of images (AP view) was generated mathematically. Images were simulated with 100k, 200k, and 300k counts for the complete image and either non- or half-perfused lesions (lesion volume 1-6 1/2% of total myocardial volume). In addition thallium images were generated excluding scatter; images were also generated with a radiopharmaceutical using 140, 184 and 296 keV, assuming the biological distribution of thallium. All images were interpreted by seven observers. Count density did not affect the true-positive fraction of the interpretations. However, the false-positive fraction decreased significantly when count density increased. Lesions located far from the camera were detected worst. The intraobserved variability was largest for interpretation of these regions. Small lesions were poorly detected. Perfusion of lesions significantly affected their detection, independent of the count density, lesion volume, and lesion location. Scatter did not affect the detection of lesions. No significant difference was found when the performance of the observers was compared for 80 keV and 140 keV. The use of 184-keV and 296-keV photons resulted in a lower detection rate probably because of the use of a high energy collimator.