Abstract
Both the theoretical and experimental data presented suggest that the diameter of the stainless steel wire used for the impulse response function should be adjusted in accordance with the spatial resolution of the CT unit being evaluated. According to theory, the standard wire size (0.356 mm) provided in the AAPM CT phantom is adequate for units with a cut-off frequency below 7 cycles/cm. In practice, this wire size could be used to test CT units with slightly higher cut-off frequencies. The pixel size used in PSF determinations should also be adjusted in accordance with the spatial resolution of the CT unit being evaluated. Experimental data suggest that pixel sizes of about one-sixth of the FWHM or 1/(5Uc) to 1/(6Uc) are most appropriate for PSF measurements. For most CT units, these recommendations translate into pixel sizes between 0.10 and 0.40 mm, depending upon the spatial resolution of the unit; CT units with the highest spatial resolution should use the smaller pixel sizes in the PSF measurements.