[Experimental and clinical studies on the accuracy of geometry-independent counter]

For the purpose to conduct precise geometry-independent radioassay of large biological samples, the authors constructed an apparatus with the "peak+scatter" counting method slightly modified from that of Gibbs, et al. and studied its accuracy from the basic and clinical aspects. The apparatus, an opposed-head counter, consisted of two matched 2 x 2-in. NaI crystal detectors and photomultipliers that were connected in parallel to a pulse-height spectrometer. The crystal faces were separated by 54 cm. The detectors were enclosed in 5cm-thick shield built with lead and the sample chamber was a 15 cubic cm-box type of 3 mm-thick Lucite. A 13 cmphiX7.7 cm styrene topper container was used for the sample container. It was found that the relation of primary gamma energy of six radioisotopes and geometry-independent counting threshold turned out to be linear. The assay accuracy with 131I was about +/- 5% in fluid samples irrespective of its sizes (up to 500 ml) or homogeneity. In solid samples the accuracy approached that of fluid by the addition of water up to the height of it. The range, in which the activity could be measured within 5% error originating from statistical fluctuations or loss of counts due to instrument dead time, proved to be about 34 approximately .06 micronCi with 57Co in assay time of 10 min. On the basis of the basic investigations, the geometry-independent counting was used for the radioassay of stool and urine. Since the assay could be done simply up to a relatively large volume without need of material adjustment, it served well in the assay of urine and stool of clinical practice as well as in animal radioassay.