Principal eigenvector field segmentation for reproducible diffusion tensor tractography of white matter structures.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Medical Image Analysis

Background:

• Diffusion tensor tractography (DTT) is crucial for mapping white matter (WM) tracts.
• Manual region of interest (ROI) selection in DTT is time-consuming and user-dependent.
• Developing automated and reproducible methods for WM tract delineation is essential.

Purpose of the Study:

• To evaluate the feasibility of an unsupervised principal eigenvector field segmentation (PEVFS) algorithm for automatic delineation of 18 major white matter tracts.
• To assess the efficiency and reproducibility of the PEVFS-based stable fiber mass (SFM) mapping approach compared to classical manual ROI methods.

Main Methods:

• Utilized diffusion tensor imaging (DTI) data to derive fractional anisotropy (FA) and principal eigenvector fields.
• Created a stable fiber mass (SFM) map by segmenting eigenvector fields using unsupervised PEVFS.
• Employed color-coded SFM segments for automatic region of interest (ROI) selection, enabling single-click fiber tractography.

Main Results:

• Successfully reconstructed all 18 targeted white matter fiber bundles in all subjects using the automated PEVFS-SFM approach.
• Demonstrated user-independent and reproducible ROI selection, enhancing the reliability of tractography.
• The PEVFS method proved robust and efficient, offering a favorable comparison to traditional manual ROI techniques.

Conclusions:

• The PEVFS-based SFM mapping provides a feasible, efficient, and reproducible automated method for white matter tract delineation in DTT.
• This approach significantly simplifies and standardizes the process of fiber tractography, reducing user dependency.
• The PEVFS method represents a promising advancement for quantitative analysis of white matter neuroanatomy.