A proposed method for correcting aperture penetration in high energy slit aperture and pinhole SPECT

Aperture penetration in high energy pinhole and slit aperture SPECT results in a decrease of both resolution and contrast. At 511 keV, approximately 50% of projection counts can be attributed to penetration. Correction for this effect by measuring the penetration from a "blocked" slit has been investigated using SPRINT-II. Slit apertures were blocked with small bars of depleted uranium (DU) and projections acquired to serve as an estimate of the penetration background. The block's finite size and DU's limited attenuation of 511 keV photons required the blocked projection data to be corrected by a spatially variant, Monte Carlo based window function. /sup 18/F line sources imaged without correction for penetration had an 8.8 mm FWHM and 18 mm FWTM due to large penetration tails. Lines reconstructed with penetration correction achieved 5.0 mm FWHM and 8.6 mm FWTM resolution. An /sup 18/F filled /spl mu/-Jaszczak phantom (hot lines from 1.5 mm to 4 mm) imaged without correction produced a barely discernible 4 mm line send uniform background. Images produced with correction produced clear images of 3 mm holes with negligible background The findings should be generally applicable to clinical gamma camera systems.