Fused multi-modal electron microscopy, a combines elastic scattering (HAADF) and inelastic spectroscopic signals (EELS/EDX) to recover high signal-to-noise ratio chemical maps at nano- and atomic-resolution.
By linking simultaneously acquired modalities through regularized optimization, the method can reduce dose requirements by over one order of magnitude while substantially improving SNR for chemical maps.
The package is available on PyPI:
pip install mapfusionFor local development, clone the repository and install in editable mode:
git clone https://github.com/jtschwar/Multi-Modal-2D-Data-Fusion.git --recursive
cd Multi-Modal-2D-Data-Fusion
pip install -e . from mapfusion import DataFusion
# Initialize fusion with list of elements
elements = ['Co', 'S', 'O']
fusion = DataFusion(elements)
# Load your chemical maps
# Provide as a dictionary where keys match your element list
# This helper function loads starter data from github
cobalt_map, sulfur_map, oxygen_map, haadf_im = fusion.load_edx_example()
chemical_maps = {
'Co': cobalt_map, # 2D numpy arrays
'S': sulfur_map,
'O': oxygen_map
}
fusion.load_chemical_maps(chemical_maps)
# Load the simultaneously acquired HAADF image
fusion.load_haadf(haadf_im) # 2D numpy array
# Run the fusion algorithm
# We can adjust with regularization parameters
fusion.run(
nIter=50,
lambdaEDS = 0.005, lambdaTV=0.1,
plot_images=True, plot_convergence=True)
# Get results in dictionary format
results = fusion.get_results()
fused_cobalt = results['Co']
fused_sulfur = results['S']Refer to the comprehensive tutorial to learn how to adjust the hyperparameters.
If you use any of the data and source codes in your publications and/or presentations, we request that you cite our papers: