Thallium-201 scintigraphy for ischaemic heart disease and infarct detection: comparison of rotating slant-hole tomography and planar imaging.

A trial was undertaken to compare the diagnostic capabilities of rotating slant-hole tomography and raw scintigraphic planar images for the assessment of ischaemic heart disease and for infarct detection. Twenty patients with coronary artery disease established by arteriography and 29 patients who had suffered acute myocardial infarction as established by ECG and enzyme criteria were imaged, along with ten normal volunteers. Overall sensitivity and specificity figures, as well as Receiver Operating Characteristics (ROC) curves, were obtained. For infarct detection there was a sensitivity of 93% for tomography and 96% for planar imaging while the specificity was 80% for tomography and 95% for planar imaging. This loss of sensitivity and specificity was also apparent in the ROC curves and showed that there was no advantage from the use of rotating slant-hole tomography. For ischaemic heart disease the sensitivities for tomography and planar imaging were 83% and 91% respectively, while the specificities were 83% and 90% respectively. The ROC analysis again showed this reduction in the diagnostic capability of the test. It was concluded that rotating slant-hole tomography was no advantage in the assessment of ischaemic heart disease. Inter-and intra-observer variability was good with consistent responses in more than 90% of cases. No significant improvement was demonstrated in the poor ability of thallium scanning to specify the diseased vessels, and the phenomenon of false defects in the rotating slant-hole tomographic scans of some normal subjects was shown to affect the response of observers significantly.