Non-negative sparse deconvolution method for PMT signals in radiation detectors

The electrical signals generated by radiation detectors are often subject to undesired effects that can affect the energy resolution of the detector. These effects may be caused by the measurement system itself or by the high rate of events. Traditional methods for shaping the signals and mitigating these effects often have limitations, such as requiring explicit information about the signals. In this paper, the authors propose a deconvolution method based on least square filtering and non-negative sparsity theory to address these limitations. The method is evaluated using simulated data representing a gamma ray detector with a NaI(Tl) crystal and a photomultiplier tube (PMT). The electrical pulses in this system are pre-processed using a CR coupling circuit and a RC integrator. The energy resolution obtained with the traditional method was 6.60%, while the proposed method resulted in an energy resolution of 5.65%. In addition to improving accuracy, the proposed method also reduces implementation complexity.