Enhanced micrometre spatial resolution and usable photon flux for hybrid pixel detectors via deep-learning-based X-ray localization
In the authors' words
Micrometre spatial resolution is achieved for X-rays with the M\"ONCH detector, a charge-integrating hybrid pixel detector with a 25 {\mu}m pixel pitch, by exploiting charge sharing between neighbouring pixels. The conventional analytical interpolation method, however, relies on a global mapping derived from the charge-sharing statistics and is applicable only to isolated single-photon clusters, which limits both the achievable resolution and the usable photon flux. To overcome these limitations, we trained deep learning models to localize single-photon and pile-up events based on high-fidelity simulation data. For knife-edge measurements at 12 keV and a low occupancy of 0.74% (photons per pixel and frame), where single-photon events dominate, the proposed method reaches a of the edge spread function of 1.36 {\mu}m, compared with 1.54 {\mu}m for the conventional eta interpolation. We demonstrate experimentally that resolving up to three photons per cluster raises the reconstruction efficiency from 59% to 93% at a 2.4{\times} higher occupancy of 1.81%, while preserving a of 1.39 {\mu}m.
Appeared: Friday, October 9. arXiv. Preprint, not yet peer-reviewed.