A statistical estimation method for counting of the prompt γ-rays from 10B(n,αγ)7Li reaction by analyzing the energy spectrum

In order to reduce the standard deviation in counting prompt γ-rays emitted from 10B(n,αγ)7Li reactions at Prompt Gamma-ray Single Photon Emission Computed Tomography (PG-SPECT) system, statistical estimation using a Maximum Likelihood Estimation (MLE) is proposed. The measured energy spectrum was analyzed statistically with MLE based on the ideal energy spectrum. The standard deviation of counting was estimated using 20 000 energy spectra including background γ-rays, generated with a Monte Carlo method on a personal computer. When the S/N ratio is lower than 0.1, the standard deviation of the method presented here, i.e. the Maximum Likelihood Estimation of Spectrum (MLES), was reduced by up to approximately 20% compared to the conventional Triple Energy Window (TEW) method.