LAUSR

Analytical electron tomography aims at achieving 3D chemical characterization at the nanoscale using energy dispersive x-ray spectroscopy (EDX), electron energy loss spectroscopy (EELS), or both simultaneously, in a scanning transmission electron microscope (STEM). Using conventional spectral processing and reconstruction algorithms, these modes require both long acquisition times, and hence electron doses, to acquire high signal-to-noise ratio spectra, and a large number of spectrum images to faithfully reconstruct the structure in 3D. To avoid damaging the sample during the tomographic acquisition, it is necessary to explore advanced machine learning methods and reconstruction algorithms [1].