[Quantification of the extent of myocardial infarction by thallium-201 single photon emission computed tomography: comparisons with post-mortem findings]

Abstract

A circumferential profile analysis with a two-dimensional polar representation of thallium(Tl)-201 myocardial single photon emission computed tomograms (SPECT) is an objective and quantitative method to evaluate myocardial infarction (MI). However, the diagnostic capability depends on the normal range of thallium distribution. Therefore, the quantitative analysis of Tl-201 myocardial SPECT was correlated with the pathological findings of the heart in 50 autopsy cases (28 with MI; 22 without MI) to determine the lower normal limits at which we can precisely determine the extent of MI on a quantitative basis. For correctly detecting the extent of MI, the lower normal limits were 2.5 standard deviations below the mean values calculated using profiles normalized to the maximum pixel count observed in each profile. Using this lower normal limit, the sensitivity was 63.8%; specificity was 87.4%; and accuracy was 80.7% for determining the extents of ischemic lesions, though these values were lower than those obtained by visual analysis. However, for detecting MI, the sensitivity by quantitative analysis was 97% which was higher than that by visual analysis. A medium sized (2.5 cm) non-transmural infarction, undetectable by visual analysis, was detectable by two-dimensional polar representations. However, the specificity of detecting MI was low (59%), and most of the false positive studies were observed in cases of valvular disease, coronary artery disease or other cardiac complications. The detectability of the extent of MI varied according to the location of the lesion. False negative lesions were most frequently observed in the septal region; false positive lesions were most frequently observed at the posterolateral regions of the dilated hypertrophic hearts. The septal region located at the peripheral portion of the MI and septal hypertrophy in cases of hypertension, were the likely negative cases, thus, these pathological characteristics may explain the high incidence of false negative results in the septal region quantitatively as well as by visual analysis. In cases with dilated hearts, the radioactivity of the thallium at the posterolateral region was more reduced by absorption compared with that in the non-dilated hearts. This reduction in counts was empirically regarded as abnormal lesions by visual analysis, however, the same situation can objectively and automatically be recognized as false positive by quantitative analysis. We concluded that quantitative analysis by Tl-201 myocardial SPECT is useful for evaluating myocardial infarction automatically, but care must be exercised concerning possible false positive diagnosis in cases with dilated hypertrophic hearts.