Data processing of radioisotope images for optimum smoothing
D C Barber, J R Mallard · Physics in Medicine and Biology · 1971
A digital method of data processing radioisotope scan images is used to derive a smoothing function for obtaining the best least mean squares fit of a noisy radioisotope image to the expected image. This is compared to smoothing with a Gaussian function, where it is seen to give a better least mean squares fit. Approximate values for the relevant parameters are derived for a digitized thyroid scan.