Anatomy-to-Tract Mapping Infers White Matter Pathways Without Diffusion Streamline Propagation.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Medical Image Analysis

Background:

• Diffusion tractography, essential for white matter mapping, is limited by diffusion MRI (dMRI) data quality issues like low signal-to-noise ratio and poor resolution.
• Existing methods often struggle with complex white matter configurations such as crossings and bottlenecks.

Purpose of the Study:

• Introduce Anatomy-to-Tract Mapping (ATM), a novel approach for generating white matter streamlines directly from T1-weighted MRI.
• To bypass the need for orientation field estimation and streamline propagation inherent in traditional dMRI tractography.

Main Methods:

• ATM utilizes the high quality and minimal distortion of anatomical MRI (T1w) to synthesize bundle-specific streamlines.
• The model is trained on multi-subject datasets, learning to condition streamline generation on subject-specific anatomy.
• Leverages global anatomical features for robust, anatomy-driven reconstructions.

Main Results:

• ATM demonstrated strong performance across multiple metrics, including bundle similarity, volume coverage, and geometric fidelity, when compared to dMRI-based methods.
• The approach successfully reconstructed 30 white matter bundles from the TractoInferno dataset.
• ATM showed superior handling of complex white matter configurations.

Conclusions:

• ATM offers a paradigm-shifting, anatomy-driven alternative to diffusion tractography for white matter pathway reconstruction.
• This method enhances structural connectivity mapping by mitigating local uncertainties present in dMRI data.
• ATM provides robust and accurate subject-specific streamline bundles by leveraging anatomical MRI.