Design and use of PET tomographs: the effect of slice spacing.

With modern positron tomographs producing 14, 21, or more transaxial slices, the effects of slice spacing on quantitative reconstruction and three-dimensional displays must be evaluated. This analysis can be approached in terms of the partial volume effect, quantified by the recovery coefficient, or in terms of sampling theory leading to the concept of aliasing. The axial recovery coefficient varies as a function of the position of an object in relation to the slices, with greater variability for larger slice spacings and finer axial resolutions. The aliased image power varies in the same way. The variability in the recovery coefficient and aliasing increase when smaller objects are imaged. Tomographs should be designed with slice spacing approximately half the full-width at half-maximum axial resolution of the tomograph; finer spacing does not appear to confer significant advantages. Thus, quantification and display in positron tomography depend on slice spacing, resolution, and object size.