Analysis of Spectrum-Imaging Datasets in Atomic-Resolution Electron Microscopy

摘要

Recently, a multivariate statistical analysis (MSA) package has been developed. In this article, the basic concept of principal component analysis (PCA), which is one of the most common MSA approaches, is described. An application of the package to an STEM atomic-column electron energy-loss spectrum-imaging dataset of Si_3N_4 is demonstrated. The PCA approach is very useful to identify statistically significant features that might be hidden under heavy random noise and to reduce efficiently random noise components. In addition, other useful auxiliary utilities available in the MSA package are briefly described.