Atomic STEM Modelling & Image Analysis
Geometry-resolved HAADF-STEM image formation and interpretable surrogate modelling with controlled validation against matched abTEM/multislice references.
- Python
- ASE
- abTEM
- NumPy
Computational Science · Atomic-Resolution Imaging
Kusse Sukuta Bersha
Computational Scientist based in Iceland.
I develop computational methods for scientific imaging and structure inference, combining physical modelling, numerical optimisation, image analysis, and reproducible scientific software.
Research
Featured research
Atomic STEM research
Current manuscripts
Paper 2 · Submission-ready unpublished manuscript
Geometry-Gated Interpretable Surrogates for HAADF-STEM Image Formation
An environment-dependent surrogate hierarchy that separates accurate one-body rendering, geometry-resolved interaction strength, model selection, and abstention.
Paper 3 · Submission-ready unpublished manuscript
Uncertainty-Aware Atomistic Inference from HAADF-STEM Images: Identifiability, Nuisance Propagation, and Prospective Experimental Qualification
A prospective inverse workflow for structural candidates, orientation, nuisance propagation, model adequacy, protected validation, and explicit abstention.
Now
Current direction
Current research focuses on geometry-resolved HAADF-STEM image formation, interpretable surrogate modelling, uncertainty-aware inference, numerical validation, and reproducible computational methods.
Background
Research background
My PhD research at the University of Iceland focused on computational reconstruction and analysis of three-dimensional atomic structures from aberration-corrected STEM data, combining image analysis, physical modelling, and numerical optimisation.
Contact
Contact
For enquiries about computational science, scientific computing, research collaboration, or related work, please use the contact page.