The Enabler Framework: an Object-Oriented Toolkit for Microscopy Data Analysis

Present-day microscopy data sets comprise numerous detector readouts that densely sample specimen properties along multiple spatial dimensions and multiple signal modalities. Typical examples include STEM EDS/EELS spectrum images, TEM tomography image stacks, and multi-channel optical sections in confocal microscopy. Such rich and complex data call for novel software tools that enable microanalysts to flexibly view, identify, extract, and quantitatively analyze their constituent signals. In the past, microscopy software development focused on the mathematical data reduction and analysis algorithms needed to isolate and quantify features or signals in digitized images and spectra [1], including filters for feature enhancement and detection, cross-correlation for drift correction and data alignment, model fitting, and various types of principal component analysis. These have been deployed within programs that run the gamut from proprietary analysis software packaged with commercial instruments to open-source software developed for general microscopy data processing (e.g. Hyperspy project However, despite the many packages, accessibility can be limited (due to high cost or to steep learning curves) and applying the tools of one package to data obtained in another can be difficult (due to proprietary data formats, incompatible abstractions, or overly specific application focus).