A mathematical approach to optimal selection of dose values in the additive dose method of EPR dosimetry

Abstract

Additive dose methods commonly used in electron paramagnetic resonance (EPR) dosimetry are time consuming and labor intensive. We have developed a mathematical approach for determining optimal spacing of applied doses and the number of spectra which should be taken at each dose level. Expected uncertainties in the data points are assumed to be normally distributed with a fixed standard deviation and the linearity of the dose response is also assumed. The optimum spacing and number of points necessary for minimal error can be estimated as can the likely error in the resulting dose estimate. When low doses are being estimated for enamel samples, the optimal spacing is shown to be a concentration of points near the zero dose value with fewer spectra taken at a single high dose value within the range of known linearity. Optimization of the analytical process results in increased accuracy and sample throughput. © 1997 Elsevier Science Ltd.