dahnke

Experimental Dahnke-style through-slice gradient correction tools for multi-echo fMRI.

The package contains:

• dahnke.correction: reusable gradient-risk and magnitude-correction functions.
• dahnke.correction.fit_complex_decay_nlls: complex R2* fitting with an optional Dahnke through-slice modulation term.
• dahnke.workflow: the motion-aware correction workflow and CLI.

The workflow can also test the combined model directly:

dahnke-motion-aware ... --output-mode both --fit-engine complex-nlls

This path estimates the Dahnke gradient, resamples native complex data as real/imaginary channels, and fits the complex decay model with the Dahnke modulation term in the forward model. With --fitmode all, the complex fit is joint per voxel across all volumes, with shared R2*/frequency and per-volume complex S0.

• docs/: Sphinx documentation, including API autodocumentation and a methodological walkthrough.
• examples/: template scripts for command-line and Python use.

Run the workflow with:

dahnke-motion-aware --help

Build the documentation with:

python -m pip install -e ".[docs]"
sphinx-build -b html docs docs/_build/html

License

dahnke is distributed under the MIT License. The package includes methods informed by the MIT-licensed t2star-mapping project; see NOTICE for the upstream copyright and license notice.