DeepTract: A Probabilistic Deep Learning Framework for White Matter Fiber Tractography

@article{Benou2019DeepTractAP,
  title={DeepTract: A Probabilistic Deep Learning Framework for White Matter Fiber Tractography},
  author={Itay Benou and Tammy Riklin-Raviv},
  journal={ArXiv},
  year={2019},
  volume={abs/1812.05129}
}
We present DeepTract, a deep-learning framework for estimating white matter fibers orientation and streamline tractography. [...] Key Method We use a recurrent neural network for mapping sequences of DWI values into probabilistic fiber orientation distributions. Based on these estimations, our model facilitates both deterministic and probabilistic streamline tractography. We quantitatively evaluate our method using the Tractometer tool, demonstrating competitive performance with state-of-the-art classical and…Expand
Track-to-Learn: A general framework for tractography with deep reinforcement learning
TLDR
This work introduces Track-To-Learn: A general framework to pose tractography as a deep reinforcement learning problem, and implements and trains algorithms to maximize returns from a reward function based on the alignment of streamlines with principal directions extracted from diffusion data. Expand
Entrack: Probabilistic Spherical Regression with Entropy Regularization for Fiber Tractography
TLDR
A general probabilistic model for spherical regression based on the Fisher-von-Mises distribution is proposed, which efficiently estimates maximum entropy posteriors of local streamline directions with machine learning methods. Expand
Track-To-Learn: A general framework for tractography with deep reinforcement learning
TLDR
This work introduces Track-To-Learn: A general framework to pose tractography as a deep reinforcement learning problem, and implements and trains algorithms to maximize returns from a reward function based on the alignment of streamlines with principal directions extracted from diffusion data. Expand
Parallel Transport Tractography
TLDR
This work proposes a novel propagation-based tracker that is capable of generating geometrically smooth curves using parallel transport frames and has a novel mechanism to reduce noise related propagation errors by incorporating topographic regularity of connections, a neuroanatomic property of many brain pathways. Expand
Deep-learning based tractography for neonates
TLDR
A convolutional neural network (CNN)-based deep-learning architecture is proposed for automated generation of WM tractography of neonates’ brain and can generally be applied for subcortical structures and cerebellum. Expand
Tractography and machine learning: Current state and open challenges
TLDR
Several datasets and evaluation tools that contain useful features for ML algorithms, along with the various methods proposed in the recent years are described, and the particular needs of ML tractography methods are described and discussed. Expand
Learning Macroscopic Brain Connectomes via Group-Sparse Factorization
TLDR
This work develops an efficient optimization strategy for this extremely high-dimensional sparse problem, by reducing the number of parameters using a greedy algorithm designed specifically for the problem, and shows that this greedy algorithm significantly improves on a standard greedy algorithm, called Orthogonal Matching Pursuit. Expand
Towards deep learning for connectome mapping: A block decomposition framework
TLDR
A new framework to map structural connectomes using deep learning and diffusion MRI and its two constituent steps can be straightforwardly incorporated as part of conventional connectome mapping pipelines to enhance accuracy is proposed. Expand
A machine learning-based method for estimating the number and orientations of major fascicles in diffusion-weighted magnetic resonance imaging
TLDR
This paper proposes a machine learning-based technique that predicts the number and orientations of crossing fascicles more accurately than several classical and machine learning methods and leads to more accurate tractography. Expand
A critical review of connectome validation studies.
TLDR
It is argued that streamline filtering can minimize spurious connections and potentially improve the biological plausibility of connectomes derived from tractography, and found that algorithmic choices such as the tractography seeding methodology, angular threshold, and streamline propagation method can substantially impact connectome reconstruction accuracy. Expand
...
1
2
...

References

SHOWING 1-10 OF 30 REFERENCES
Learn to Track: Deep Learning for Tractography
TLDR
This work uses feed-forward and recurrent neural networks to learn the generation process of streamlines directly from diffusion-weighted imaging data and suggests that exploiting past information within a streamline's trajectory during tracking helps predict the following direction. Expand
Direct Estimation of Fiber Orientations Using Deep Learning in Diffusion Imaging
TLDR
This work presents a novel approach for estimating the fiber orientation directly from raw data, by converting the model fitting process into a classification problem based on a convolutional Deep Neural Network, which is able to identify correlated diffusion information within a single voxel. Expand
Fiber tractography using machine learning
TLDR
A Fiber tractography approach based on a random forest classification and voting process, guiding each step of the streamline progression by directly processing raw diffusion-weighted signal intensities demonstrates the vast potential of machine learning for fiber tractography. Expand
Fiber tractography using machine learning
TLDR
A Fiber tractography approach based on a random forest classification and voting process, guiding each step of the streamline progression by directly processing raw diffusion-weighted signal intensities demonstrates the vast potential of machine learning for fiber tractography. Expand
White matter fiber tractography: why we need to move beyond DTI.
TLDR
This comprehensive study shows that the most widely used clinical tractography method results in systematically unreliable and clinically misleading information, and needs to move beyond the diffusion tensor framework to begin to provide neurosurgeons with biologically reliable tractography information. Expand
A Machine Learning Based Approach to Fiber Tractography Using Classifier Voting
TLDR
The presented machine learning based approach to fiber tractography is the first of its kind and its experiments showed auspicious performance compared to 12 established state of the art tractography pipelines. Expand
Data-driven fiber tractography with neural networks
TLDR
This work presents a purely data-driven neural network regression model for fiber tractography, which benefits from additional data, which can be incorporated easily, even from different supervisors, and is robust to noise and variation in the data. Expand
Combining white matter diffusion and geometry for tract-specific alignment and variability analysis
TLDR
A framework for along-tract analysis of white matter (WM) fiber bundles based on diffusion tensor imaging (DTI) and tractography and the construction of an FFDD dissimilarity measure for sub-voxel alignment of fiber bundles, based on the fast marching method (FMM), is presented. Expand
Probabilistic streamline q-ball tractography using the residual bootstrap
TLDR
The proposed residual bootstrap method utilizes a spherical harmonic representation for high angular resolution diffusion imaging (HARDI) data in order to estimate the uncertainty in multimodal q-ball reconstructions. Expand
Reconstruction of Diffusion Anisotropies Using 3D Deep Convolutional Neural Networks in Diffusion Imaging
TLDR
This work presents a new approach to derive the fiber orientation distribution function using a Deep Convolutional Neural Network, which remains stable, even if less gradient directions are acquired. Expand
...
1
2
3
...