Quantification of technetium 99m-labeled sestamibi single-photon emission computed tomography based on mean counts improves accuracy for assessment of relative regional myocardial blood flow: experimental validation in a canine model.

BACKGROUND

Quantification of single-photon emission computed tomographic (SPECT) images is generally based on determination of maximal counts on radial sectors of short-axis slices. We hypothesized that analysis of mean counts may reduce estimation error.

METHODS AND RESULTS

We compared quantitative 99mTc-labeled sestamibi (MIBI) SPECT based on maximal myocardial counts with that based on mean myocardial counts for accuracy of quantifying relative regional myocardial perfusion in a canine model of permanent left anterior descending coronary artery occlusion. MIBI and radiolabeled microspheres were injected during left anterior descending coronary artery occlusion. Relative microsphere myocardial blood flow was expressed as a percentage of normal (left circumflex coronary artery territory) blood flow. SPECT imaging was performed in vivo and ex vivo. Relative MIBI uptake on SPECT short-axis slices was quantified with normalized circumferential profiles based on maximal and mean counts. In vivo and ex vivo SPECT relative myocardial count density was compared to relative myocardial blood flow in six dogs. In the comparisons, percent errors in estimating the relative blood flow and relative flow deficit with MIBI SPECT imaging were calculated. There was an excellent correlation between absolute myocardial tissue MIBI activity and regional myocardial blood flow for each of the six dogs (r = 0.90 to 0.98). The correlations between relative myocardial count density on SPECT and relative blood flow for individual sectors were similar for maximal and mean count profiles (maximal, 0.79 to 0.83; mean, 0.77 to 0.82). Comparing the nadirs of in vivo and ex vivo circumferential count profiles, the correlations were slightly better (maximal, 0.82 to 0.91; mean, 0.87 to 0.91). Average percent errors in assessing relative blood flow and relative flow deficit were decreased significantly by use of mean count profiles (p < 0.05).

CONCLUSIONS

Relative SPECT count density with either maximal or mean count profiles correlated well with relative myocardial blood flow. Compared with maximal count profiles, quantification with mean count profiles improved estimation of relative flow.