Time-coded aperture design for nuclear medicine imaging: a study of signal-to-noise ratio

A signal-to-noise analysis is presented for time varying, pseudorandomly modulated coded apertures. An expression for optimal code transmission is derived and evaluated as a function of sampling density for a source model simulating a heart with varying liver background. Optimal performance is obtained for dense sampling and mean code transmission between 20 and 30%. In the absence of liver background, optimal coded aperture efficiency is equal to that of the ideal nonmultiplexed aperture. Adding a liver background equal to 2.2 times the object activity effectively lowers the coded aperture efficiency a factor of 2. However, practical constraints on the ideal nonmultiplexing aperture reduce its efficiency a factor of 4.