Knesaurek K
Division of Nuclear Medicine, University of Rochester Medical Center, New York 14642.
Med Phys. 1992 May-Jun;19(3):591-8. doi: 10.1118/1.596850.
Two new experimental methods of calculating scatter fraction (SF), as a function of depth, are presented and used in a comparison study. These methods are based on the assumptions that the total point spread function (psft) consists of geometrical (psfg) and scattering (psfs) components, and that the psft and its components are radially symmetric functions. Both methods assume that psfg is a two-dimensional (2-D) Gaussian function. The first proposed method is based on the serial model of the psft. According to the serial model, psfs is also a 2-D Gaussian function. However, the second method is based on the assumption that the psfs is a monoexponential function. Thus the main difference between these two experimental methods is the way in which the psfs is treated. Published data obtained by Monte Carlo simulations compared to the new experimental results shows that for depths less than 5 cm, both new methods yield smaller values for scatter fraction than Monte Carlo simulation or the subtraction method. However, for depths from 5 to 20 cm, the results of these two new methods were very close to the values obtained by both the Monte Carlo simulation and the subtraction method. These two methods have allowed assessment of the scatter fraction with higher accuracy and reliability and reporting data concludes that the new methods are less error sensitive than the subtraction method.
