Asymmetric Fiber Orientation Distribution Functions
Asymmetric Fiber Orientation Distribution Functions (AFOD)
Asymmetric Fiber Orientation Distribution Functions (AFODs) are advanced representations in diffusion MRI (dMRI) used to estimate the orientation of white matter fibers in the brain. Unlike traditional symmetric Fiber Orientation Distribution Functions (FODs), AFODs capture asymmetric sub-voxel fiber configurations, such as bends, branching, and junctions, which are common in complex brain anatomy.
Background
FODs measure the directional diffusion of water molecules in biological tissues. Conventional FOD estimation methods, such as spherical deconvolution, assume antipodal symmetry, treating directions +v and -v identically.[1]. While effective in simple fiber configurations, this assumption loses critical anatomical details in complex regions, motivating the development of AFODs [2].
Asymmetric FODs (AFODs)
AFODs address the limitations of symmetric FODs by capturing true asymmetry in fiber orientation. Techniques to achieve AFOD estimation include:
- Neighbourhood-based constrained spherical deconvolution, which leverages information from adjacent voxels to infer asymmetry [3].
- Cone-model filtering, which applies directional spatial filters to emphasize asymmetric fiber populations [4].
- Asymmetric FOD approaches such as the Asymmetric Response Function (ARF) method, which adjusts the fiber response function for asymmetry [5].
- Unsupervised deep learning methods for AFOD estimation that learn asymmetric representations from diffusion MRI data [6].
These methods allow more biologically meaningful representation of white matter architecture, enhancing downstream applications such as tractography and connectivity analysis [2][4][5][6].
Machine Learning Approaches
Deep learning models have been applied to estimate FODs directly from diffusion signals. These models learn complex, non-linear mappings and can integrate spatial context to improve accuracy[7]. However, many still assume symmetry or rely on supervised data generation, limiting their ability to detect true fiber asymmetries [7][8][9].
Encoder-based Curvature-Aware Regularization (EnCAR)
Encoder-based Curvature-Aware Regularization (EnCAR) is a novel method designed to improve AFOD estimation using a curvature-aware encoder with geometric regularization [9].
EnCAR improves estimation in complex regions while maintaining performance in symmetric regions. It integrates with AFOD pipelines, enhancing tractography and connectivity analyses by providing anatomically accurate fiber orientations [9].
Potential Applications
- Improved tractography through more accurate fiber trajectories in regions with bends and junctions [2][4][9].
- Enhanced quantitative connectivity analysis by better representing microstructural complexity [2][10]
- Clinical studies, including detection of white matter abnormalities in neurodegenerative diseases and brain injuries.[4]
Summary
- FODs are central to diffusion MRI analysis but standard symmetric models ignore anatomical asymmetry.[1][2]
- AFODs address asymmetry but are limited by traditional estimation methods.[2][10]
- Well-known asymmetric approaches include ARF and unsupervised deep learning methods.[5][6]
- asymmetric methods have the potential to transform tractography, connectivity mapping, and clinical white matter analysis by providing richer, geometrically faithful representations of brain microstructure.[5][6][9][4][2]
References
- ↑ 1.0 1.1 Tournier, J-Donald; Calamante, Fernando; Connelly, Alan (2007-05-01). "Robust determination of the fibre orientation distribution in diffusion MRI: Non-negativity constrained super-resolved spherical deconvolution". NeuroImage. 35 (4): 1459–1472. doi:10.1016/j.neuroimage.2007.02.016. ISSN 1053-8119. PMID 17379540.
- ↑ 2.0 2.1 2.2 2.3 2.4 2.5 2.6 Basser, Peter J.; Pajevic, Sinisa; Pierpaoli, Carlo; Duda, Jeffrey; Aldroubi, Akram (October 2000). "In vivo fiber tractography using DT-MRI data". Magnetic Resonance in Medicine. 44 (4): 625–632. doi:10.1002/1522-2594(200010)44:4<625::AID-MRM17>3.0.CO;2-O. ISSN 0740-3194. PMID 11025519.
- ↑ Jeurissen, Ben; Tournier, Jacques-Donald; Dhollander, Thijs; Connelly, Alan; Sijbers, Jan (2014-12-01). "Multi-tissue constrained spherical deconvolution for improved analysis of multi-shell diffusion MRI data". NeuroImage. 103: 411–426. doi:10.1016/j.neuroimage.2014.07.061. ISSN 1053-8119. PMID 25109526.
- ↑ 4.0 4.1 4.2 4.3 4.4 Tournier, J.-Donald; Calamante, Fernando; Connelly, Alan (December 2013). "Determination of the appropriate b value and number of gradient directions for high-angular-resolution diffusion-weighted imaging". NMR in Biomedicine. 26 (12): 1775–1786. doi:10.1002/nbm.3017. ISSN 0952-3480. PMID 24038308.
- ↑ 5.0 5.1 5.2 5.3 Bastiani, Matteo; Cottaar, Michiel; Dikranian, Krikor; Ghosh, Aurobrata; Zhang, Hui; Alexander, Daniel C.; Behrens, Timothy E.; Jbabdi, Saad; Sotiropoulos, Stamatios N. (2017-09-01). "Improved tractography using asymmetric fibre orientation distributions". NeuroImage. 158: 205–218. doi:10.1016/j.neuroimage.2017.06.050. ISSN 1053-8119. PMC 6318223. PMID 28669902.
- ↑ 6.0 6.1 6.2 6.3 Zhang, Di; Li, Ziyu; Deng, Xiaofeng; Han, Zekun; Wang, Alan; Liu, Yong; Zong, Fangrong (2026-05-01). "Asymmetric fiber orientation distribution estimation via unsupervised deep learning". Medical Image Analysis. 110. doi:10.1016/j.media.2026.103968. ISSN 1361-8415. PMID 41702180 Check
|pmid=value (help). Unknown parameter|article-number=ignored (help) - ↑ 7.0 7.1 Yang, Yiqiong; Yuan, Yitian; Ren, Baoxing; Wu, Ye; Feng, Yanqiu; Zhang, Xinyuan (2025-03-05). "Deep Learning-Based Diffusion MRI Tractography: Integrating Spatial and Anatomical Information". NeuroImage. 317. arXiv:2503.03329. doi:10.1016/j.neuroimage.2025.121314. PMID 40570535 Check
|pmid=value (help). Unknown parameter|article-number=ignored (help) - ↑ Karimi, Davood; Vasung, Lana; Jaimes, Camilo; Machado-Rivas, Fedel; Warfield, Simon K.; Gholipour, Ali (2021-10-01). "Learning to estimate the fiber orientation distribution function from diffusion-weighted MRI". NeuroImage. 239. doi:10.1016/j.neuroimage.2021.118316. ISSN 1053-8119. PMC 8385546 Check
|pmc=value (help). PMID 34182101 Check|pmid=value (help). Unknown parameter|article-number=ignored (help) - ↑ 9.0 9.1 9.2 9.3 9.4 Taherkhani, Mojtaba; Pizzolato, Marco; Morup, Morten; Dyrby, Tim B. (2026-04-02). "Encoder-based Curvature-Aware Regularization for estimating asymmetric fiber orientation distribution functions in diffusion MRI". bioRxiv 10.64898/2026.03.31.715534 Check
|biorxiv=value (help). - ↑ 10.0 10.1 Dell'Acqua, Flavio; Rizzo, Giovanna; Scifo, Paola; Clarke, Rafael Alonso; Scotti, Giuseppe; Fazio, Ferruccio (March 2007). "A Model-Based Deconvolution Approach to Solve Fiber Crossing in Diffusion-Weighted MR Imaging". IEEE Transactions on Biomedical Engineering. 54 (3): 462–472. Bibcode:2007ITBE...54..462D. doi:10.1109/TBME.2006.888830. ISSN 0018-9294. PMID 17355058.
This article "Asymmetric Fiber Orientation Distribution Functions" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Asymmetric Fiber Orientation Distribution Functions. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
