The Determination of Body Burdens of Radionuclides by Computer Analysis of Gamma-ray Spectral Data

Publisher Website
Google Scholar
Abstract
The technique of y-ray measurement by pulse-height analysis provides the most rapid and accurate determination of internally-deposited radionuclide abundances in man. Scintillation spectrometer pulse-height distribution data were analyzed to quantify the levels of 137Cs, 65Zn, 40K and other radionuclides in 500 subjects. In a large-scale study of this kind, it was found desirable to employ automatic data-handling techniques, including computer analysis of the spectral data. The whole-body counter data of the subjects were compared with that obtained from a phantom (REMAB) containing standardized solutions of a number of monoenergetic gamma emitters. A number of Fortran II programs for the IBM-7094 were used for the reduction of the pulse-height distribution data. The programs first perform a non-linear least-squares fit of a Gaussian function to the data points in the spectrum photopeaks. The associated Compton distribution required in the stripping program is then computed by interpolation from an experimentally-derived matrix of seven monoenergetic gamma rays. Commencing with the highest energy, each designated photopeak and its associated Compton is subtracted off until the complex spectrum is analyzed. Other programs take the initial spectral data from the analyzer in the form of punched paper tape and convert it to magnetic tape for use in the above stripping programs. This technique described appears capable of providing rapid and accurate quantitative interpretation of pulse-height distribution data, applicable to the determination of body burdens of various radionuclides. It will be shown that the method is less subjective than the usual graphic method for interpreting spectral data. The accuracy attainable and limitations of this computer analysis will be discussed along with the results of specific population surveys for estimating body burdens.

This publication has 0 references indexed in Scilit: