Big Data Analytics for Scanning Transmission Electron Microscopy Ptychography

Abstract

Electron microscopy is undergoing a transition; from the model of producing only a few micrographs, through the current state where many images and spectra can be digitally recorded, to a new mode where very large volumes of data (movies, ptychographic and multi-dimensional series) can be rapidly obtained. Here, we discuss the application of so-called “big-data” methods to high dimensional microscopy data, using unsupervised multivariate statistical techniques, in order to explore salient image features in a specific example of BiFeO3 domains. Remarkably, k-means clustering reveals domain differentiation despite the fact that the algorithm is purely statistical in nature and does not require any prior information regarding the material, any coexisting phases, or any differentiating structures. While this is a somewhat trivial case, this example signifies the extraction of useful physical and structural information without any prior bias regarding the sample or the instrumental modality. Further interpretation of these types of results may still require human intervention. However, the open nature of this algorithm and its wide availability, enable broad collaborations and exploratory work necessary to enable efficient data analysis in electron microscopy.

In‐situ Transmission Electron Microscope Techniques for Heterogeneous Catalysis

“…However, considering the large sizes of TEM images, the high‐speed recording cameras could create several gigabytes of data per minute or even per second, which brings a great challenge to laboratory engineers and scientists for data storage and processing. To efficiently process the massive volume of data and extract maximum information from the available data, researchers, therefore, need more efficient solutions for data storage and processing in terms of both computer hardware and software …”

Next-Generation Information Technology Systems for Fast Detectors in Electron Microscopy

“…Irrespective of their application, using fast electron detectors at their full rates generates large amounts of raw data. This situation creates significant challenges for data handling and processing (Belianinov, et al, 2015) (Jesse, et al, 2016). One option is to immediately reduce the data in a suitable fashion and to handle only a reduced data stream.…”

USID and Pycroscopy – Open Source Frameworks for Storing and Analyzing Imaging and Spectroscopy Data

“…4 shows the use of the SVD and Cluster classes in Pycroscopy to extract trends in 4-D STEM datasets that are invisible to the human eye. 112 Typically, the raw Convergent Beam Electron Diffraction (CBED) patterns (2D images) recorded by the camera detector are averaged at each position of the electron beam to provide a pair of 'bright-field' and 'dark-field' 2D images.…”