AN INVESTIGATION OF FILTER CHOICE FOR FILTERED BACK-PROJECTION RECONSTRUCTION IN PET

A key parameter in the practical application of filtered back-projecti on (FBP), the standard clinical image reconstruction algorithm for pos itron emission tomography (PET), is the choice of a low-pass filter wi ndow function and its cut-off frequency. However, the filter windows a nd cut-off frequencies for clinical reconstruction are usually chosen empirically, based on a small sample of images and filters. By conside ring the features of the signal and noise spectra in a sinogram, the d esired image resolution, and the signal-to-noise ratio (SNR) of the fi ltered sinogram, a methodology for informed selection of a filter func tion and cut-off frequency for FBP was investigated. Simulations of si nogram data similar to whole body or cardiac studies provided informat ion on the signal and noise frequency-domain spectra of noisy projecti on data. The improvements in SNR with different filter windows and cut -off frequencies were evaluated and compared. The projection spectrum SNR measure did not prove to be an accurate indicator of subjective im age quality or lesion detectability with variations in Poisson noise a nd image resolution.