Evaluation of scatter compensation methods by their effects on parameter estimation from SPECT projections

Three algorithms for scatter compensation in Tc-99m brain single-photon emi ssion computed tomography (SPECT) were optimized and compared on the basis of the accuracy and precision with which lesion and background activity cou ld be simultaneously estimated. These performance metrics are directly rela ted to the clinically important tasks of activity quantitation and lesion d etection, in contrast to measures based solely on the fidelity of image pix el values. The scatter compensation algorithms were (a) the Compton-window (CW) method with a 20% photopeak window, a 92-126 keV scatter window, and a n optimized "k-factor," (b) the triple-energy window (TEW) method, with opt imized widths of the photopeak window and the abutting scatter window, and (c) a general spectral (GS) method using seventeen 4 keV windows with optim ized energy weights. Each method was optimized by minimizing the sum of the mean-squared errors (MSE) of the estimates of lesion and background activi ty concentrations. The accuracy and precision of activity estimates were th en determined for lesions of different size, location, and contrast, as wel l as for a mon complex Bayesian estimation task in which lesion size was al so estimated. For the TEW and GS methods, parameters optimized for the esti mation task differed significantly from those optimized for global normaliz ed pixel MSE. For optimal estimation, the CW bias of activity estimates was larger and varied more (-2% to 22%) with lesion location and size than tha t of the other methods. The magnitude of the TEW bias was less than 7% acro ss most conditions, although its precision was worse than that of CW estima tes. The GS method performed best, with bias generally less than 4% and the lowest variance; its root-mean square (rms) estimation error was within a few percent of that achievable from primary photons alone. For brain SPECT, estimation performance with an optimized, energy-based, subtractive correc tion may approach that of an ideal scatter-rejection procedure. (C) 2001 Am erican Association of Physicists in Medicine.