An iterative computer algorithm for optimization of radionuclide subtraction studies.

A new algorithm for off-line computer optimization of radionuclide subtraction studies is presented. Within the digitized image, any region of interest may be chosen for study. No prior knowledge is required concerning the position and extent of the principal and interfering objects, nor concerning their image intensities. Tests show that even when statistical fluctuations are relatively large, the intensity quotient calculated by this algorithm may be used to estimate a uniform background and correct for its contribution. The effect of smoothing the images prior to the use of the algorithm has been investigated. The method has also been used for pancreas imaging and one example of this application is presented.