Accuracy of fluorod dose measurements.

Abstract

Aminiature glass rod dosimeter (fluorod) has recently been described as a useful and reliable radiation detector for a variety of medical dosimetry problems (1–5). The percentages of accuracy cited by the various workers, however, for their dose measurements appear to be estimates rather than experimentally determined values. In the course of our own studies (6, 7) it became obvious that an experimental determination of the limits of accuracy obtainable with fluorods was mandatory. Experimental Procedure A large number of new fluorods,2 from 71 to 137 at a time, were exposed simultaneously to a homogeneous 2-Mvp x-ray beam. The fluorods were placed in a large water phantom, 5 cm. below the surface of the water, at the center of a 20 × 20-cm. field at a TSD of 100 cm. This method of exposure assured an equal dose to all fluorods in each set. Six sets of fluorods were so irradiated with measured doses ranging from 155 to 2,795 rads (Table I). Prior to their irradiation, the background (inherent) fluorescence of each fluorod was measured with a fluorimeter described previously (6). Only fluorods with background readings equivalent to between 60 and 100 rads of 2-Mvp x-rays (3 to 5 divisions on the fluorimeter scale) were used in this study. An average background value for the fluorods of each set was determined and, after the irradiation, was subtracted from the individual fluorescence reading of each fluorod. Since the background value was an average of the range of 60 to 100 rads, this introduced an uncertainty of about ±20 rads in the dose measurements. Because the radiation-induced fluorescence increases for several hours after exposure (8), the fluorescence readings were made twenty-four to forty-eight hours after irradiation. Special precautions were taken to insure a high degree of cleanliness of the fluorods at the time of their measurement, since surface contamination may yield additional fluorescence. Each fluorod was washed in a strong soap solution and rinsed successively in water, acetone, and 95 per cent alcohol. The fluorod was then dried in air and wiped with lint-free lens paper. Results Examples of the experimentally observed dose distributions obtained with each set of irradiated fluorods are shown in Figures 1 and 2. The dose measurements are expressed in fluorimeter scale divisions. Each scale division represents 20.5 rads. Superimposed on the scatter diagram is the normal distribution curve based on the experimental data. The probability that the distribution of the measured values differs significantly from the normal distribution was calculated by means of the chi-square test. Table II presents the chi-square analysis of the data shown in Figure 2. The data are distributed in 10 intervals. The chi-square value is 6.01. The corresponding probability value for 10 intervals is approximately 0.80 (9).