3-D smoothing techniques for dynamic scintigraphic images

A new 3-D Double Window Adaptive Hybrid (3-D DWAH) smoothing algorithm that smooths data in both space and time domains was applied to dynamic renal scintigraphic images. Anscombe and inverse Anscombe transformations were used before and after filtering to make Poisson noise signal independent. Then, the renogram, R(t), taken from a filtered region of interest (ROI over the kidney) and the input function, B(t), (ROI over the heart) are deconvolved for obtaining the renal retention function, H(t), representing the spectrum of transit times through the kidney. This has been used to calculate reliable values of mean transit time (MTT) and glomerular filtration rate (GFR). The images of both MTT and GFR parameters effectively display the transit and uptake information available from the renogram. The results obtained by using the 3-D DWAH smoothing algorithm were compared by the results of other conventional smoothing techniques and unsmoothed raw data.<>