New nDTomo software from our Research & Development Lead Scientist Dr Antony Vamvakeros
We are pleased to announce the launch of new nDTomo software from our Research & Development Lead Scientist, Dr Antony Vamvakeros. You can find nDTomo available on PyPI.
nDTomo is an open-source Python software suite for simulation, visualisation, reconstruction, and analysis of chemical imaging and X-ray tomography data. It is especially useful for hyperspectral datasets like XRD-CT.
This software has been seven years in the making. Antony started the project during his postdoctoral at the ESRF – The European Synchrotron building GUIs for handling XRD-CT data at beamline ID15A. Over the years he has developed nDTomo to be far from a GUI – now itβs an integrated platform for researchers and students in materials science, catalysis, batteries, and synchrotron applications.
nDTomo features include:
π Interactive visualization of chemical tomography data via the nDTomoGUI
π§ͺ Generation of multi-dimensional synthetic phantoms
π― Simulation of pencil beam CT acquisition strategies
π§Ό Pre-processing and correction of sinograms
π οΈ CT image reconstruction using algorithms like filtered back-projection and SIRT
π§ Dimensionality reduction and clustering for unsupervised chemical phase analysis
π Pixel-wise peak fitting using Gaussian, Lorentzian, and Pseudo-Voigt models
π€ Peak fitting using the self-supervised PeakFitCNN
π Simultaneous peak fitting and tomographic reconstruction using the DLSR approach with PyTorch GPU acceleration
Antony worked on this software alongside Finden colleagues; Dr Evangelos Papoutsellis and Dr Hongyang Dong.
nDTomo is a helpful new tool if you are working with XRD-CT, chemical tomography, or hyperspectral imaging, so try it for yourself.
More information at:
π Docs: https://ndtomo.readthedocs.io/en/master/
β GitHub: https://github.com/antonyvam/nDTomo
The new software release includes:
β’ π Full API documentation + 10 Jupyter notebooks, including recent work we have done at the TLDR group at Dyson School of Design Engineering with Ronan Docherty and Prof Sam Cooper on a self-supervised neural network for peak fitting
β’ π§ Transition from tensorflow to PyTorch for all neural-network and GPU-based tools (e.g. PeakFitCNN, DLSR)
β’ π§ͺ Major GUI upgrades, including XRD-CT phantom generator + Embedded IPython console
β’ π§Ή Refactored, cleaned, and simplified codebase
β’ π First stable release to PyPI