Noise reduction in PET attenuation correction using non-linear Gaussian filters

In a PET study, shortening of transmission scan time is highly desired for improving patient comfort and increasing scanner throughput. It necessitates a method that reduces statistical noise in attenuation correction factors. The authors have evaluated non-linear Gaussian (NLG) filtering for smoothing transmission images reconstructed with filtered backprojection. In contrast to a conventional linear Gaussian filter, the weights of the NLG filter depend not only on difference in pixel locations but also on difference in attenuation values. Several filtering steps with different NLG parameters can effectively reduce noise without losing structural information. In this method, there is no need to reconstruct transmission images with iterative algorithms, and no assumption of attenuation coefficients used in segmentation methods. Results with phantom and patient data suggested that the NLG filtering method is useful for attenuation correction using count-limited transmission data for both brain and whole-body PET studies