Fitting and benchmarking of dosimetry data for new brachytherapy sources.

New source designs of encapsulated low-energy gamma emitting nuclides for permanent implants require dosimetric analysis and calibration standardization. The dosimetry measurements can be incorporated into a treatment planning system by fitting the data. The use of a fitting function whose behavior at range limits mimics the physical phenomena, using as few parameters as possible, eliminates noisy outliers and lends credence to calculations beyond the measured range. Clinical implementation of the new sources also requires benchmarking against existing sources, where the current clinical experience lies. We present an analysis of measured dosimetry data for three brachytherapy sources recently available from North American Scientific, Inc. (North Hollywood, CA): Pd-103 source model MED3633 ("PdGold"), I-125 source models MED3631-A/M ("IoGold-AM") and MED3631-A/S ("IoGold-AS"). Using the formalism of the Interstitial Collaborative Working Group (ICWG) the radial dose function, g(r), the anisotropy function, F(r, theta), and the anisotropy factor, phi (an)(r), were previously evaluated from measurements of each source design. In this report we use fitting functions whose forms are chosen to approach reasonable values at data limits. These forms are quite similar to those used in a previous analysis of TG43 Iodine and palladium compendium data. Fitting parameter results for each function are provided for each brachytherapy source model. Fit-data discrepancies are smaller than measurement uncertainties, meaning that incorporation into treatment planning systems will not introduce significant errors in clinical use. Current clinical experience is based on the Theragenics (Norcross, GA) Pd-103 seed ("PdThera"), and the Nycomed-Amersham (Arlington Heights, IL) I-125 seed models 6711 ("6711") and 6702 ("6702"). The new sources are benchmarked against these seeds. (C) 2000 American Association of Physicists in Medicine. [S0094-2405(00)01410-3].