Liom Toolkit#
LIOM Toolkit is a Python package for processing and analyzing light-sheet fluorescence microscopy (LSFM) data. It supports the Laboratoire d’Imagerie Optique et Moléculaire at Polytechnique Montréal and is published to PyPI for the broader neuroimaging community.
The toolkit covers the full LSFM workflow: format conversion
(HDF5/NIfTI/NRRD → OME-Zarr), brain registration to the Allen Atlas (ANTsPy),
vessel segmentation (classical + PyTorch U-Net), and morphometric statistics.
Other labs can pip install liom-toolkit and run the complete
conversion → registration → segmentation → stats pipeline without source
edits or hardcoded lab config.
Installation, first run, and the optional ai / antspy / all
extras for deep learning and atlas registration.
Auto-generated reference for the 7 liom-* command-line tools,
rendered from each script’s _build_argument_parser().
Auto-generated reference for the liom_toolkit Python package,
produced by sphinx-autoapi from static AST analysis.
Worked examples: conversion, segmentation, templating, and stats — code-only (run locally with all extras for outputs).
Prose overview of the conversion → registration → segmentation → stats pipeline and how the pieces fit together.
Dev environment, tests, lint/typing, and the correctness rules every change to this package must respect.