Bragg curve identification using a neural network

Bragg curve spectrometers (BCS) used as heavy ion detectors, have been very useful in the study of nuclear reactions for a number of years. A novel way to process the anodic signal (BC) from a BCS is to digitize this signal using a waveform recorder. However, it poses a problem in terms of how to process this large amount of information. One of the options is to use neural networks. For this purpose, it is convenient to consider the determination of the total energy parameter (E) and the Bragg peak height (BP) from a BC as a pattern recognition problem. Neural network have demonstrated to be a very good option to address this kind of task. In a preliminary study, a neural network was used to identify a single one parameter (BP or E) from a BC. In this paper we report the results obtained for a simultaneous identification of both parameters.