Hierarchical Unsupervised Clustering of Nanobeam Electron Diffraction Data Using Sparse Bragg-Vector Representations
Abstract
If disk detection is applied to nanobeam electron diffraction datasets, then the results are effectively a list of vectors describing the position of every diffraction peak in real and reciprocal space. This is the natural territory for the application of clustering algorithms, and they are shown to be highly effective at decomposing such datasets and automating imaging and analysis. In particular, this is found to work particularly well in a hierarchical clustering approach, where a first level is concerned with finding clusters of similar diffraction vectors and a second level with grouping those diffraction vector clusters that originate from the same real space area (i.e. crystal or crystallographic variant). Examples are shown in both polycrystalline and single crystal (with precipitates) systems. Additionally, automated separation of amorphous or deeply nanocrystalline components is also found to be possible in an optional third level of clustering, allowing composite images of both amorphous and crystalline components in partially crystallised samples to be easily and automatically generated. These advances promise to increase throughput in atomic structure analysis with nanobeam diffraction, and also make finding minor components much easier. They can also serve as a preliminary step towards more detailed crystallographic or crystal size/shape distribution analysis.
Related
- Orientation refinement for nanobeam precession electron diffraction patterns using spot intensity profiles
- Bliss of dimensionality: How an unsupervised criterion identifies optimal low-resolution representations of high-dimensional datasets
- Coarse-Graining Hidden Representations: Unsupervised Neuron Selection via Mapping Entropy
- Probabilistic, Structure-Intrinsic Clustering with an Hierarchical Embedding (PSICHE) in Time and Space Applied In Particle Physics Jet Reconstruction
- Compact thermo-optic switch based on graphene-silicon nitride nanobeam cavity
BibTeX
@misc{maclaren2026hierarchical,
title = {{Hierarchical Unsupervised Clustering of Nanobeam Electron Diffraction Data Using Sparse Bragg-Vector Representations}},
author = {Ian MacLaren and Ala Al-Afeef and Trevor Almeida and Rantej Bali and Shriyar Tariq and Emily Wackan and Luke Daly and Joshua F. Einsle},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2606.23201},
eprint = {2606.23201},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.23201},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Hierarchical Unsupervised Clustering of Nanobeam Electron Diffraction Data Using Sparse Bragg-Vector Representations
AU - Ian MacLaren
AU - Ala Al-Afeef
AU - Trevor Almeida
AU - Rantej Bali
AU - Shriyar Tariq
AU - Emily Wackan
AU - Luke Daly
AU - Joshua F. Einsle
PY - 2026
DA - 2026/10/07
PB - arXiv
DO - 10.48550/arxiv.2606.23201
UR - https://arxiv.org/abs/2606.23201
AB - If disk detection is applied to nanobeam electron diffraction datasets, then the results are effectively a list of vectors describing the position of every diffraction peak in real and reciprocal space. This is the natural territory for the application of clustering algorithms, and they are shown to be highly effective at decomposing such datasets and automating imaging and analysis. In particular, this is found to work particularly well in a hierarchical clustering approach, where a first level is concerned with finding clusters of similar diffraction vectors and a second level with grouping those diffraction vector clusters that originate from the same real space area (i.e. crystal or crystallographic variant). Examples are shown in both polycrystalline and single crystal (with precipitates) systems. Additionally, automated separation of amorphous or deeply nanocrystalline components is also found to be possible in an optional third level of clustering, allowing composite images of both amorphous and crystalline components in partially crystallised samples to be easily and automatically generated. These advances promise to increase throughput in atomic structure analysis with nanobeam diffraction, and also make finding minor components much easier. They can also serve as a preliminary step towards more detailed crystallographic or crystal size/shape distribution analysis.
ER -