Neural-based pile-up correction and ballistic deficit correction of X-ray semiconductor detectors using the Monte Carlo simulation and the Ramo theorem

Pile-up distortion is a common problem in many nuclear radiation detection systems, especially in high count rates. It can be solved by hardware-based pile-up rejections, but there is no complete pile-up elimination in this way. Additionally, the methods can lead to poor quantitative results. Generally, time characteristics of semiconductor detector pulses are different from Scintillator detector pulses due to ballistic deficit. Hence, pulse processing-based pile-up correction in the detectors should consider this specification. In this paper, the artificial neural network pile-up correction method is applied for silicon detector piled-up pulses. For this purpose, the interaction of photons with a silicon detector is simulated by the MCNP4c code and the pulse current is calculated by Ramo's theorem. In this approach, we use a sub-Nyquist frequency sampling. The results show that the proposed method is reliable for pile-up correction and ballistic deficit in semiconductor detectors. The technique i...