AI-assisted Data Analytical Tool for Secondary Electron Hyperspectral Imaging (in Matlab)

This repository provides AI-assisted Data Analytical Tool for Secondary Electron Hyperspectral Imaging (SEHI) and Scanning Electron Microscope (SEM) derived Secondary Electron (SE) spectroscopy. It is Matlab implementation of machine learning approach presented in the paper ''Towards Automated Chemical Analysis of Materials Using Secondary Electron Hyperspectral Imaging and Unsupervised Learning'', which can be used for high-throughput chemical analysis of materials at the micro- and nano-scale.

This work is supported by the UK EPSRC EP/V012126/1 the grant ‘‘SEE MORE, MAKE MORE: Secondary Electron Energy Measurement Optimisation for Reliable Manufacturing of Key Materials’’. Contact: SM3 (SEE MORE MAKE MORE) project PI, Professor Cornelia Rodenburg, c.rodenburg@shefield.ac.uk.

We also acknowledge the support from Insigneo Institute for In Silico Medicine in Sheffield.

This repository contains the Matlab source code of this proposed approach, ReadMe file, license, and example data [1] for demo. For code details and user guidance, please see the ReadMe file in the folder. This given example data is a processed SEHI image stack of printed Metal alloy and Carbon film (Palladium Silver Carbon complex film). More example data is given from our database [1]. More relevant details about the example can also be found from the paper [2].

Please note: this Matlab code package requires Matlab R2019a or latter installed. It also requires Image Processing Toolbox (for reading image in TIF files) and Statistics and Machine Learning Toolbox (Gaussian mixture model). Part of the code (Dirichlet Process Gaussian Mixture Model) is based on the public MATLAB package available from GitHub [3] by Mo Chen (https://github.com/PRML/PRMLT.git), released under MIT license.

Regarding the data pre-processing, we have previously released two tools, one is python implementation [4], and the other is the software [5]. This Matlab code package is compatible with the open-source python tool [4]. Please check the available tools depending on your needs.

If you find any issues or have suggestions, please contact with Jingqiong Zhang (jingqiong.zhang@sheffield.ac.uk). Please cite this repository inline with MIT license when using the code in your work.

[1] J. Zhang, J. Nohl, N. Farr, L. Mihaylova, C. Rodenburg, K. Abrams, K. Black, ‘‘SEHI (Secondary Electron Hyperspectral Imaging) dataset of Metal alloy and Carbon film (Palladium Silver Carbon complex film),’’ The University of Sheffield, Dataset, https://doi.org/10.15131/shef.data.22202923.v1, 2023.

[2] K. Abrams, M. Dapor, N. Stehling, M. Azzolini, S. Kyle, J. Schäfer, A. Quade, F. Mika, S. Kratky, Z. Pokorna, I. Konvalina, D. Mehta, K. Black, C. Rodenburg, ‘‘Making Sense of Complex Carbon and Metal/Carbon Systems by Secondary Electron Hyperspectral Imaging,’’ Advanced Science, https://doi.org/10.1002/advs.201900719, 2019.

[3] C. Mo, ‘‘Prml:pattern recognition and machine learning,’’ https://github.com/PRML/PRMLT.git, 2019.

[4] J. Nohl, ‘‘pysehi releases version 1.1.1,’’ The University of Sheffield, Software, https://doi.org/10.15131/shef.data.22310068, 2023.

[5] J. Nohl, N. Farr, N. Stehling, J. Zhang, F. Longman, G. Ives, L. Mihaylova, C. Holland, C. Rodenburg,‘‘Csehi app 1.0,’’ https://doi.org/10.15131/shef.data.21647090.v1, 2023.