Comparison of methods of attenuation and scatter correction in brain perfusion SPECT.

OBJECTIVE The radioactivity count distribution in the brain must be determined accurately to accurately measure cerebral blood flow (CBF). Scatter and attenuation are factors that compromise the accuracy of determining radioactivity counts in the brain. METHODS We compared regional CBF in patients by an autoradiographic method using N-isopropyl-p-[123I] iodoamphetamine when, first, attenuation correction alone was performed uniformly on SPECT images by using empiric mu-values (Chang method); second, scatter correction was performed and the mu-values of a homogeneous-attenuation body of water were used for attenuation correction (triple-energy window [TEW]+Chang method); and third, scatter correction was performed and the mu-values calculated by CT were used for attenuation correction (TEW+CT method). We also compared regional CBF measured by these methods with the values obtained by the xenon CT/CBF method, which uses CT and stable xenon. RESULTS Scatter correction reduced overestimation of regional CBF in low-flow regions. The TEW+CT method yielded better regional and overall correlations with the xenon CT/CBF method than did either of the other methods. CONCLUSION The TEW+CT method of correction gave the most accurate measurements of regional CBF.