Fiber-driven resolution enhancement of diffusion-weighted images

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Computational Neuroscience

Background:

• Diffusion-weighted imaging (DWI) provides valuable insights into brain circuitry but suffers from limited spatial resolution and signal-to-noise ratio (SNR).
• Current DWI techniques often struggle to accurately resolve complex fiber structures and assess brain connectivity due to inherent resolution limitations.

Purpose of the Study:

• To develop and evaluate an algorithm for increasing the resolution of DWI beyond the original scan resolution.
• To enable a more detailed investigation of white matter fiber structures and improve the accuracy of brain connectivity assessments.

Main Methods:

• The algorithm generates a dense vector-valued field by interpolating diffusion data along axonal fibers, guided by the fiber orientation distribution function (ODF).
• Interpolation weights are assigned based on the likelihood of fiber traversal in specific directions, ensuring data is sourced from aligned fibers.
• This directional weighting preserves local fiber structures and mitigates blurring caused by averaging data from misaligned fibers.

Main Results:

• The proposed algorithm significantly reduces blocking artifacts in the enhanced images.
• It achieves greater smoothness in anatomical structures compared to conventional methods.
• Markedly improved visibility of structural details and fiber pathways was observed.

Conclusions:

• The developed algorithm effectively enhances DWI resolution, offering a superior method for investigating brain circuitry.
• This technique holds promise for more accurate and detailed assessments of brain connectivity and white matter integrity.