Probabilistic tractography using Q-ball imaging and particle filtering: Application to adult and in-utero fetal brain studies

Abstract

By assuming that orientation information of brain white matter fibers can be inferred from Diffusion-Weighted Magnetic Resonance Imaging (DW-MRI) measurements, tractography algorithms provide an estimation of the brain connectivity in vivo. The two key ingredients of tractography are the diffusion model (tensor, high-order tensor, Q-ball, etc.) and the means to deal with uncertainty during the tracking process (deterministic vs probabilistic mathematical framework). In this paper, we investigate the use of an analytical Q-ball model for the diffusion data within a well-formalized particle filtering framework. The proposed method is validated and compared to other tracking algorithms on the MICCAI'09 contest Fiber Cup phantom. Tractographies of in vivo adult and fetal brain Diffusion-Weighted Images (DWIs) are also shown to illustrate the robustness of the algorithm.

Brain white matter fiber estimation and tractography using Q-ball imaging and Bayesian MODEL

“…are widespread throughout the human brain [11], and traditional tensor models cannot adequately handle their fiber bundle geometry [3]. Therefore, the HARDI framework was proposed to deal with these issues.…”

“…The presented method used spherical deconvolution [18] and normalization [19] to improve the accuracy and robustness of the Q-ball model. Further details about QBI and ODF can be found in [3] and [20].…”

“…These indirect observations of white matter geometry allow nerve fiber reconstruction through the use of tractography algorithms [3]. The reconstruction of white matter fibers in the human brain can be used to investigate connectivity in mental and neurological disorders.…”

“…To overcome these problems, the global optimal framework for brain tractography was proposed [6][7][8][9][10]. Global parameters were optimized using the diffusion data and some constraints to get global fiber bundles [3]. However, the global optimal framework for brain tractoraphy is computationally expensive because an optimized estimation must be made for the entirety of the fiber bundles.…”

“…However, the global optimal framework for brain tractoraphy is computationally expensive because an optimized estimation must be made for the entirety of the fiber bundles. Another way to overcome the issue of noise and partial volume effect in the diffusion data is to use a probabilistic framework, which characterizes the uncertainty in fiber path estimation with a set of possible propagation directions for a given voxel [3]. Probabilistic methods can generate multiple fiber trajectories with different probabilities and a probability map from the seed voxel to the other voxels in the brain [11][12][13].…”