Difference between revisions of "Projects:DiffusionGraphBasedConnectivity"
| Line 10: | Line 10: | ||
[[Image:UNC_GraphbasedConnectivity_Ex1.png|thumb|300px|Slicer visualization of genu connectivity map overlaid on streamline genu tracts]] | [[Image:UNC_GraphbasedConnectivity_Ex1.png|thumb|300px|Slicer visualization of genu connectivity map overlaid on streamline genu tracts]] | ||
Regional connectivity measurements derived from diffusion imaging datasets are of considerable interest in the neuroimaging community for better understanding white matter connectivity. Current connectivity measurements are usually either based on fiber tractography applied in a Monte-Carlo fashion or are variations of the Hamilton-Jacobi approach. | Regional connectivity measurements derived from diffusion imaging datasets are of considerable interest in the neuroimaging community for better understanding white matter connectivity. Current connectivity measurements are usually either based on fiber tractography applied in a Monte-Carlo fashion or are variations of the Hamilton-Jacobi approach. | ||
| + | |||
| + | [[Image:UNC_GraphbasedConnectivity_Monkeycombined.png|thumb|300px|Average cost map computed on the rhesus monkey DWI atlas with the source located in the internal capsule (left) and the genu of the corpus callosum (right). The grayscale overlays show the FA map, and the 3D visualizations include the streamline tractography results from the same seed regions for comparison purposes.]] | ||
We propose a novel, graph-based algorithm that provides a fully deterministic, efficient and stable connectivity measure. This method handles crossing fibers and deals well with multiple seed regions. The computation is based on a multi-directional graph propagation algorithm applied to sampled orientation distribution functions computed directly from the original diffusion imaging data. A maximum probability and a minimum cost approach are both possible. | We propose a novel, graph-based algorithm that provides a fully deterministic, efficient and stable connectivity measure. This method handles crossing fibers and deals well with multiple seed regions. The computation is based on a multi-directional graph propagation algorithm applied to sampled orientation distribution functions computed directly from the original diffusion imaging data. A maximum probability and a minimum cost approach are both possible. | ||
| Line 16: | Line 18: | ||
We are currently working on the application of our algorithm to a publicly available synthetic dataset generated via Numerical Fiber Generator (NFG) (Close et al. A software tool to generate simulated white matter structures for the assessment of fibre-tracking algorithms. NeuroImage (2009) vol. 47 (4) pp. 1288-300). This will allow us to thoroughly evaluate the algorithm as well as compare it to various existing connectivity/tractography methods. | We are currently working on the application of our algorithm to a publicly available synthetic dataset generated via Numerical Fiber Generator (NFG) (Close et al. A software tool to generate simulated white matter structures for the assessment of fibre-tracking algorithms. NeuroImage (2009) vol. 47 (4) pp. 1288-300). This will allow us to thoroughly evaluate the algorithm as well as compare it to various existing connectivity/tractography methods. | ||
| − | |||
| − | |||
| − | |||
[[Image:UNC_connectivity_nfg.png|thumb|300px|Application to NFG synthetic data. We start with a complex synthetic dataset, and we compute connectivity maps for each ROI defined by the start/end regions of each fiber bundle. These individual cost maps are compiled into a connectivity matrix that represents the connectivity values between each pair of ROI's. The ROI pairs along the diagonal (marked with a white dot) are the true connections (as given by the ground truth) in the data.]] | [[Image:UNC_connectivity_nfg.png|thumb|300px|Application to NFG synthetic data. We start with a complex synthetic dataset, and we compute connectivity maps for each ROI defined by the start/end regions of each fiber bundle. These individual cost maps are compiled into a connectivity matrix that represents the connectivity values between each pair of ROI's. The ROI pairs along the diagonal (marked with a white dot) are the true connections (as given by the ground truth) in the data.]] | ||
Revision as of 17:33, 17 October 2011
Home < Projects:DiffusionGraphBasedConnectivityBack to UNC Algorithms
Muti-directional graph propagation for Diffusion Imaging based Connectivity
This project focuses on connectivity measurements derived from diffusion imaging datasets in order to better understand cortical and subcortical white matter connectivity. Our research employs a novel, multi-directional graph propagation method that performs a fully deterministic, efficient and stable connectivity computation. The method handles crossing fibers and deals well with multiple seed regions. In addition to the analysis of these connectivity measures in describing brain pathology, they can also be used as scalar maps for use in DTI registration.
Description
Regional connectivity measurements derived from diffusion imaging datasets are of considerable interest in the neuroimaging community for better understanding white matter connectivity. Current connectivity measurements are usually either based on fiber tractography applied in a Monte-Carlo fashion or are variations of the Hamilton-Jacobi approach.
We propose a novel, graph-based algorithm that provides a fully deterministic, efficient and stable connectivity measure. This method handles crossing fibers and deals well with multiple seed regions. The computation is based on a multi-directional graph propagation algorithm applied to sampled orientation distribution functions computed directly from the original diffusion imaging data. A maximum probability and a minimum cost approach are both possible.
So far we have results on very simple synthetic and limited real datasets to illustrate the potential of our method towards subject-specific connectivity measurements performed in an efficient, stable and reproducible manner. Such individual connectivity measurements would be well suited for studies of neuropathology.
We are currently working on the application of our algorithm to a publicly available synthetic dataset generated via Numerical Fiber Generator (NFG) (Close et al. A software tool to generate simulated white matter structures for the assessment of fibre-tracking algorithms. NeuroImage (2009) vol. 47 (4) pp. 1288-300). This will allow us to thoroughly evaluate the algorithm as well as compare it to various existing connectivity/tractography methods.
Publications
Key Investigators
- UNC Algorithms: Ipek Oguz, Wenyu Lu, Alexis Boucharin, Yundi Shi, Martin Styner
