One framework to understand the workflow of much of experimental science, including electron microscopy of materials, is a progression from data to information to knowledge. Data are the more-or-less raw numbers measured in experiments. In electron microscopy, data include images, spectra, and diffraction patterns. Information is facts about the object of study derived from the data. In microscopy, information includes atomic positions and elemental identities and available electron energy levels. Knowledge is the answer to scientific questions and the ultimate goal of research.
This talk will review examples of and opportunities for computational microscopy at every step along the chain. Data can be improved by removing noise, correcting distortions and blurring, and reducing electron dose and damage to the sample. Information can be obtained from data more precisely, effectively, and efficiently using signal unmixing, deep learning, or autonomous experiments. Even knowledge may be obtained by machine learning of physical laws from patterns in information. Constraints and prior information from the microscope instrument, electron scattering physics, and materials structure will be highlighted throughout.
Back to Computational Microscopy Tutorials