bonndit
The bonndit package contains computational tools for diffusion MRI processing developed at the University of Bonn.
bonndit implements constrained single and multi tissue deconvolution with higher-order tensor fODFs [Ankele17], and the extraction of principal fiber directions with low-rank tensor approximation [Schultz08]. It also includes code for fiber tractography based on higher-order tensor fODFs, and for filtering the resulting set of streamlines. In particular, bonndit implements spatially regularized tracking using joint tensor decomposition or an Unscented Kalman Filter [Gruen23]. It also contains code from a study in which we compared the strategy of selecting the most suitable number of fiber compartments per voxel to an adaptive model averaging which reduced the model uncertainty [Gruen22].
Finally, the package includes code for suitably constrained fitting of the Diffusional Kurtosis (DKI) model, and computation of corresponding invariants [Groeschel16].
Free software: GNU General Public License v3
Documentation: https://bonndit.readthedocs.io.
Installation
To install bonndit via pip, run the following command
$ pip install bonnditTo install bonndit via conda, run
$ conda install bonndit -c xderes -c conda-forgeFeatures
An overview of the scripts and functionality included in bonndit is given in our documentation. It also includes a tutorial for performing fiber tracking with our code.
Reference
If you use our software as part of a scientific project, please cite the corresponding publications. The method implemented in stdeconv and mtdeconv was first introduced in
Michael Ankele, Lek-Heng Lim, Samuel Groeschel, Thomas Schultz: Fast and Accurate Multi-Tissue Deconvolution Using SHORE and H-psd Tensors. In: Proc. Medical Image Analysis and Computer-Aided Intervention (MICCAI) Part III, pp. 502-510, vol. 9902 of LNCS, Springer, 2016
It was refined and extended in
Michael Ankele, Lek-Heng Lim, Samuel Groeschel, Thomas Schultz: Versatile, Robust, and Efficient Tractography With Constrained Higher-Order Tensor fODFs. In: Int’l J. of Computer Assisted Radiology and Surgery, 12(8):1257-1270, 2017
The methods implemented in low-rank-k-approx was first introduced in
Thomas Schultz, Hans-Peter Seidel: Estimating Crossing Fibers: A Tensor Decomposition Approach. In: IEEE Transactions on Visualization and Computer Graphics, 14(6):1635-42, 2008
The methods implemented in peak-modelling was first introduced in
Johannes Grün, Gemma van der Voort, Thomas Schultz: Reducing Model Uncertainty in Crossing Fiber Tractography. In proceedings of EG Workshop on Visual Computing for Biology and Medicine, pages 55-64, 2021
Extended in:
Johannes Grün, Gemma van der Voort, Thomas Schultz: Model Averaging and Bootstrap Consensus Based Uncertainty Reduction in Diffusion MRI Tractography. In: Computer Graphics Forum 42(1):217-230, 2023
The regularized tractography methods (joint low-rank and low-rank UKF) were first implemented in prob-tracking and introduced in
Johannes Grün, Samuel Gröschel, Thomas Schultz: Spatially Regularized Low-Rank Tensor Approximation for Accurate and Fast Tractography. In NeuroImage 271:120004, 2023
The use of quadratic cone programming to make the kurtosis fit more stable which is implemented in kurtosis has been explained in the methods section of
Samuel Groeschel, G. E. Hagberg, T. Schultz, D. Z. Balla, U. Klose, T.-K. Hauser, T. Nägele, O. Bieri, T. Prasloski, A. MacKay, I. Krägeloh-Mann, K. Scheffler: Assessing white matter microstructure in brain regions with different myelin architecture using MRI. In: PLOS ONE 11(11):e0167274, 2016
PDFs can be obtained from the respective publisher, or the academic homepage of Thomas Schultz: https://cg.cs.uni-bonn.de/person/prof-dr-thomas-schultz
Credits
This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.
History
0.3.1 (2023-03-15)
Included UKF tractography
Included regularized tractography
0.2.0 (2021-09-17)
Included the missing steps of the whole tracking pipeline.
0.1.2 (2019-02-26)
‘mtdeconv’: If response is available, files needed for the computation of the response are not loaded.
0.1.1 (2019-02-06)
First release on PyPI.
0.1.0 (2019-02-06)
Making repository public on Github
2 maintainers
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
File details
Details for the file bonndit-0.3.1.tar.gz.
File metadata
- Download URL: bonndit-0.3.1.tar.gz
- Upload date:
- Size: 331.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via:
twine/4.0.2 CPython/3.9.16
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b8296939a66d18e298aafb885d101922d2c6a9083a2cf51ea4127964576b6d73
|
|
| MD5 |
89a72b9260accc818c7349ee7c0c7de9
|
|
| BLAKE2b-256 |
86f353316ae704e0b3f791972319575d0463206d748e01e42c97d92670f5739c
|
