Simulation Study on Unfolding Methods for Diagnostic X-rays and Mixed Gamma Rays

A photon detector operating in current mode that can sense X-ray energy distribution has been reported. This detector consists of a row of several segment detectors. The energy distribution is derived using an unfolding technique. In this paper, comparisons of the unfolding techniques among error reduction, spectrum surveillance, and neural network methods are discussed through simulation studies on the detection of diagnostic X-rays and gamma rays emitted by a mixture of 137Cs and 60Co. For diagnostic X-ray measurement, the spectrum surveillance and neural network methods appeared promising, while the error reduction method yielded poor results. However, in the case of measuring mixtures of gamma rays, the error reduction method was both sufficient and effective.