Class separation and parameter estimation with neural networks for the XEUS wide field imager

The X-ray space telescope XEUS is the proposed follow-up project to ESAʹs cornerstone mission XMM-Newton which is now in orbit. To face the high data rate from the pixel detector and to improve event processing neural networks are under study to be integrated into the electronics on board (online) and to serve as analysis tool on ground (offline). For two applications results are presented. First as a typical online application, the separation of single photon events from pileup: here the unwanted event topologies are separated from useful ones belonging to a single X-ray photon. Second a typical off-line application, the estimation of the incident position of a photon: here the charge splitting (i.e. signal charges are collected by two or more adjacent pixels) can be used to determine a precise incident position of a photon. The neural network results are compared with standard methods.