µMatch: 3D shape correspondence for microscopy data

µMatch (microMatch) is a shape correspondense library for a variety of mesh types.

Installation

• Clone the repository.
• Create a conda virtual environment and install the package
• Install the project using the source

git clone <this repo url>
cd muMatch
conda env create -f environment.yml
conda activate mumatch
pip install .

Installation for development mode

• Clone the repository.
• Create a conda virtual environment and install the package
• Install the project using the source in editable mode

git clone <this repo url>
cd muMatch
conda env create -f environment.yml
conda activate mumatch
pip install --editable .

Getting started

Once the package has been installed the pipeline can be tested with:

from micro_match.pipeline import run_micromatch_test

run_micromatch_test()

This analysis pipeline runs all steps of the microMatch pipeline on the example data we provide in the example_data folder and can be used as a basis to use microMatch on your own data. It also illustrates some basic shape analysis outputs that can be retreived once correspondence has been established.

Feel free to adjust the pipeline for your needs in the "micro_match/run_pipeline.py" file.

We also provide a further example of visualization of the correspondence map in the jupyter notebook scalar_plotting.ipynb. Once run_pipeline.py has been run once, scalar_plotting.ipynb can be explored in your browser through the jupyter notebook interface by running

jupyter notebook

with the muMatch conda environment activated (see Installation above).

Troubleshooting

• If you adapt run_pipeline.py directly to work on your own files, make sure that your file names do not include full stops, dashes or underscores (the shorter the name, the less risk of problems).
• It is possible to transfer data from Imaris (.wrl) to meshlab (conversion from .wrl to .stl) and use .stl in muMatch.

Optional: automatic house-keeping

To maintain a good code base, you can setup pre-commit hooks to:

• format your code (using a code formatter like black)
• inspect your code for potential caveats (using a linter like flake8)
• canonically sort your import (using isort)

To setup pre-commit setup, you can type when your environment is activated:

pip install pre-commit black flake8 isort
pre-commit install

How to cite

If you use µMatch, please cite us as follows:

J. Klatzow, G. Dalmasso, N. Martinez-Abadias, J. Sharpe, and V. Uhlmann, "μMatch: 3D shape correspondence for microscopy data", Frontiers in Computer Science, 2022. https://doi.org/10.3389/fcomp.2022.777615