Deep Normative Tractometry for Identifying Joint White Matter Macro- and Micro-structural Abnormalities in...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biomedical Engineering

Background:

• Brain white matter integrity is crucial for cognitive function.
• Neurodegenerative diseases like Alzheimer's disease cause white matter microstructural alterations.
• Current methods for analyzing white matter abnormalities have limitations in fine-scale mapping.

Purpose of the Study:

• To introduce the Deep Normative Tractometry (DNT) framework for analyzing brain white matter.
• To encode joint macrostructural and microstructural profiles of white matter tracts.
• To detect and map along-tract abnormalities in neurodegenerative conditions.

Main Methods:

• Developed a variational autoencoder (VAE) based framework (DNT).
• Trained DNT on healthy control data to learn normative tract distributions.
• Assessed DNT's generalizability using transfer learning on an independent cohort.
• Compared DNT results with the Bundle Analytics (BUAN) tractometry pipeline.

Main Results:

• DNT successfully delineated along-tract micro- and macro-structural abnormalities.
• Detected widespread diffusivity abnormalities in mild cognitive impairment and Alzheimer's disease cohorts.
• DNT findings closely aligned with established BUAN pipeline results.
• Demonstrated potential to distinguish anisotropy abnormalities from tract macrostructure.

Conclusions:

• DNT provides a robust framework for analyzing white matter tract integrity.
• The framework shows promise for enhancing fine-scale mapping of white matter alterations.
• DNT has significant potential for improving the detection of neurodegenerative disease-related changes.