BAYESIAN APPROACH TO AUTOMATIC MASS SPECTRUM PEAK IDENTIFICATION IN ATOM PROBE TOMOGRAPHY

摘要

Atom probe tomography relies on the analysis of mass-spectrometry data [1, 2] and requires preliminary identification of the mass-spectrum peaks. Establishing correct correspondence between the peaks in the spectrum and the particular atomic species is crucial for accurate interpretation of the measured data [3, 4]. Manual peak labeling is a time-consuming procedure and vulnerable to errors [4]. We propose a Bayesian approach suitable for robust automatic peak identification.