Adaptively constrained convex optimization for accurate fiber orientation estimation with high order spherical...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biomedical Engineering

Background:

• Human Connectome Project (HCP) diffusion imaging offers high-resolution in vivo whole-brain white matter mapping.
• Accurate reconstruction of fiber orientation distribution is crucial for understanding brain connectivity.

Purpose of the Study:

• To develop a novel method for accurate fiber orientation distribution reconstruction from diffusion imaging data.
• To reliably resolve crossing fibers with small separation angles using advanced techniques.
• To apply the method to state-of-the-art HCP data for studying complex fiber structures.

Main Methods:

• Developed a novel method solving the spherical deconvolution problem as a constrained convex optimization problem.
• Employed adaptively selected constraints and high-order spherical harmonics.
• Validated the method on simulated data and multi-shell/diffusion spectrum imaging (DSI) data from HCP.

Main Results:

• The proposed algorithm demonstrated superior performance in resolving fiber crossings compared to a popular spherical deconvolution method on simulated data.
• Successfully applied the method to HCP multi-shell and DSI data.
• Showcased the method's capability in analyzing intricate white matter fiber structures.

Conclusions:

• The novel constrained convex optimization approach reliably reconstructs fiber orientation distribution from diffusion imaging.
• This method enhances the study of complex white matter architecture using advanced neuroimaging data.
• The technique holds promise for improved in vivo brain connectivity mapping.