Robust Topographical Representation for Longitudinal Propagation of Tau Pathology

Area of Science:

• Neuroscience
• Computational Biology
• Medical Imaging

Background:

• Tau pathology is a key indicator of Alzheimer's disease (AD) progression.
• Longitudinal tau positron emission tomography (PET) offers insights into disease evolution.
• Current computational models struggle with spatial and temporal data from tau PET.

Purpose of the Study:

• To develop a novel representation for spatiotemporal tau PET data.
• To improve computational modeling of Alzheimer's disease progression.
• To enhance the analysis of longitudinal tau PET data for AD research.

Main Methods:

• Proposed a Multiresolutional Reeb Graph representation for tau topography.
• Developed topology-based similarity and severity interleaving distance metrics.
• Validated the representation in subtyping, staging, and longitudinal prediction tasks.

Main Results:

• The Multiresolutional Reeb Graph effectively encodes spatiotemporal tau propagation.
• The proposed metrics enable robust quantification of pathology distribution and longitudinal staging.
• Downstream tasks showed improved performance compared to existing methods.

Conclusions:

• The novel representation enhances the application of longitudinal tau PET data in AD research.
• Offers a reliable computational approach for studying Alzheimer's disease progression.
• Demonstrates significant potential for advancing AD biomarker analysis.