he duplicated. There is considerably less variety in sments of 1. Oral Biol. choice of materials employed. While rice, sugar, or F .imilar granular materials are frequently used to simu- Inte the uncontaminated body, it is the almost uni- versal custom to employ aqueous solutions whenradioiJ. Microseogk .«livity is to be incorporated therein. I’or intensive and 'adiat. Res. ried use, a phantom containing radioactive liquid F ..- several practical disadvantages, namely the everF present danger of leakage, which maylead to contamim of Calei F uation of expensively shielded detectors and enclosure, 153 (1963), and the inconvenience of cross contamination between ill, J. H. Co - different isotopes used within the phantom itself. The of Cats Iutter counsels the use of short-lived materials, and the n Annual RE additional labor that their use entails. These problems 189, pp. 24-38 increase in importance with the complexity in detail tion of the Ig i the phantom. Biologicalk Less hazardous and tedious manipulation may be ratory Radié abtained by incorporating the desired radioelement ort, Januan ynne Nation into beads of suitable size. These may be poured from y, J. Radiuk ~tornge container to phantom manytimes with quan- ’ titative transfer. While they are still subject to accidental spillage, recovery and salvage are considerably m in Hume, simplified. We have recently investigated the pracBone. Radio, tirnlity of such a system, and present here a progress w-Hill, Nei report on some of its advantages and problems. Appearan | A.J., Schol, itive Micry seratic Bone PREPARATION OF BEADS ‘:.dioactive microspheres of either ceramic or plastic material are available from the 3M Companyin sizes ringing from 10 to 150 micrometers.“ On special 4J., Arnstein® . é ycline-B ase, order, they supplied us with much larger plastic spheres, te and Cel 725 + 70 in diameter. Specific activities normally Orthoped. 46° produced range from 0.1 to 30 Ci/g. These activity levels -re not compatible with our intended use, but, again the Alkaline bs special request, we obtained spheres loaded with resium-137 and eobalt-60 at approximately 1.3 wCi/g. Institute o The spheres are much too small to handle individS10, uly or to locate if accidentally spilled (about 5000/g); therefore, we obtained nylon beads 5.5mm diameter » with a I-mm hole drilled to the approximate center of Fenech. (The indentation will hold about 20 microspheres,) A fixed number of spheres were placed in cach eavity and sealed in place with a drop of low\isvosity epoxy cement. A vacuum pickup device, udjusted to accept a fixed number of spheres, devised by Mr, E. Fudala of the Radiological Physics Division stuff shop, greatly simplified the loading procedure. ) simulate STANDARDIZATION Whenliquid sources are used, a uniform distribution automatically results. With discrete beads, one must either (1) assay each bead separately, identify it uniquely, and sum the total activity in a given source, or (2) determine the mean activity level of the beads and their distribution about that value. We chose the latter course for obvious reasons. One hundred microspheres were counted repetitively (a different sample each time) with a total of 58 samples; a few additional runs were made with only 10 or 20 microspheres, i.e., single beads. Table 31 summarizes the results. Data are expressed as net counts per minute in the entire photopeak, as detected by a 17.8-cm diameter by 8.9-cm thick Nal crystal and single channel ana- lyzer. Sources were placed in accurately reproducible geometryon the crystal face. The first loading of beads was carried out using microspheres from the top of the container. The abso- lute activity level was about 30% greater than that inferred from the assay supplied by the 3M Company. Wespeculated that perhaps the discrepancy might be due to nonuniform labeling of the plastic resin spheres, which were then insufficiently intermixed and ran- domized during subsequent processing and shipping. Hence we sampled spheres taken from the midplane and bottom of the container and noted significant differences, though nothing like 30% (see lines 4 and 5 Table 31). The large values of chi squared indicated that the samples tested do not represent a normal distribution. We observe that many spheres are in actuality more like half-spheres, with a smaller fraction of intermediate shapes. It has been suggested that allowing the spheres to roll down a shghtly inclined plane would separate those with large flat surfaces and hence improvethe uniformity, and we intend to trythis. With this activity level, filling a liver or kidneyentirely with source beads results in an inconveniently large total activity. The remedyis to use the required number of active beads, diluted with additional nonradioactive ones. For economy, the latter may be of Lucite instead of nylon. As a practical matter, we think it likely that not more than 100 active spheres will usually suffice for a given requirement. The number actually used maywell be determined bythestatistical distribution in space required to simulate a uniform e purporf Migure 57 illustrates beads held in a simple polyethylene distribution, rather than by the precision in total f varying, Cich lot was sprayed with paint of a distinctive color to e physica, “‘cird against accidental admixture and to facilitate dy are té ' !rleval if inadvertently dropped. 0.7 g/cm. For accurate stmulation of the larger organs, where self-absorption may be appreciable, fine granules could be added to increase the density to unity. ion equip? jig while being loaded and sealed. Beads were loaded With each isotope at two specific activity levels, and activity. The mean density of close-packed beads is about