Do patient data ever exceed the partial volume limit in gated SPECT studies?

BACKGROUND

Some single photon emission computed tomography (SPECT) methods to detect percent myocardial wall thickening (%WT) assume a linear relationship to changes in maximum myocardial counts, predicated on myocardial walls never exceeding the SPECT camera's partial volume limit. Recent studies have challenged such assumptions, reporting that systolic count changes underestimate wall thickening as measured by echocardiography and magnetic resonance imaging.

METHODS AND RESULTS

To test whether clinical data ever are observed to exceed the partial volume limit, we examined gated tomograms of 75 patients selected at random and of an additional 25 patients known to have hypertension with electrocardiographic evidence of left ventricular hypertrophy. Image transformations were performed such that for every cinematic frame, radial counts at every angle were automatically normalized to the same maximum count. If no patient's myocardium ever exceeded the partial volume limit, thickness quantified from transformed images would always be the same throughout the cardiac cycle and would just correspond to the camera's line spread function. Thickness was measured by Gaussian fitting of transformed myocardial counts in the epicardial direction only to exclude cavitary count contamination. % WT was computed from thickness differences from diastole to systole. % WT values were assessed from clinical data at lateral, inferior, septal, anterior, and apical territories. Resulting %WT distributions were tested against the null hypothesis of %WT = 0 by the Z-test. Although some distributions were not actually Gaussian, the maximum mean %WT was only +3% +/-5% for the septal wall, in agreement with an observer's impressions of no detectable wall thickening. Thus mean %WT values were trivial compared with expected physiologic normal values of 30% to 50%.

CONCLUSION

No convincing evidence was found of thickness above the partial volume limit in this large sample of 75 normotensive and 25 hypertensive patients. Therefore it is likely that relations between myocardial count increases and wall thickening are similar throughout the cardiac cycle, even in patients with left ventricular hypertrophy.