80 FOR: a t "22" _ 4# fa “a a 40 cm = B= ——_—__—_—_+] ko, |+-_— 20em Net CITTTE etal 0.25 0.! = 10 (&,=450 key} 4 Re = 3.361 x I | Ri, = 2.299 F =i,462 RA, = 3.026 Fo o= ft.ill R, = 3.272 ~ | 1.027 a) [ I 1 l | x | | (2) [ Il t I ] wt me t q 1 Fig. 60—Comparison of single, double, and quadrupole point source approximations to a uniform line source f, thus obtained, for a choice of parameters typical of many counting systems, is compared with the response of one-, two-, and four-point sources having equal total intensity and distributed symmetrically. It is found that two sources result in a response which is 11% low, three sources (not illustrated)—d %, and four sources—2.7 %. There is one big drawback. The point source nearest the detector contributes 56% of the total response for four sources, and the first two contribute 82%. They must, therefore, be located accurately at the designated places, but of course in many phantom systemsthis is impractical, if not impossible. On the other hand, one might envision a phantom system in which a uniform distribution within any selected organ, or indeed any arbitrary activity distribution, could be simulated by a set of equal point sources located at uniquely determined points of a three-dimensional grid or coordinate system. A computer program permitting a systematic investigation of organ size, shape and composition based upon such a concept has been reported by Snyderet al. We are inclined to believe that a more useful approach is to determine the most probable response and the statistical distribution about the mean resulting from successively large numbers of points locai random. One mighttry to do this empirically by p a sufficient number of beads in a suitable enc] shaking them up thoroughly, and observing tl sultant count rate. The question of what constit “sufficient number, and what is the distribution tion would then be evaluated from a large num! repeated trials. A more sophisticated approach involve Monte Carlo simulation by computer. In gation of the properties of an ensemble of sc along these lines is continuing. In conclusion, an alternate system for the safe ing of small amounts of radioactivity into phat has been described. We feel that the method has improved flexibility (being capable of simulating about anyarbitrarily chosen source configuration safety. Further analysis of the random spatial «i bution assumed by multiple sources, and the resu detector response in realistic, three-dimensional « dinates is necessary in order to realize its full pote: REFERENCES 1. Grotenhuis, I. M. Properties and Uses of Radiating 4 spheres. Radioactive Pharmaceuticals, Ed. G. A. An+