Quantitative analysis of dynamic tau brain positron emission tomography (PET) imaging using cluster initialized factor analysis

This study investigates the application of Cluster Initialized Factor Analysis (CIFA) to enhance the quantitative analysis of dynamic tau brain PET imaging utilizing ¹⁸F-Flurotaucipir (FTP). We analyzed PET scans from 18 subjects (9 Pittsburgh Compound-B (PIB)-positive and 9 PIB-negative) to uncover distinct tracer-binding dynamics, which describe the temporal patterns of radiotracer uptake in brain regions. Specific binding refers to the interaction of the radiotracer with its intended target, such as tau proteins in Alzheimer’s disease. By examining factor curves resulting from the dynamic image factorization and comparing the corresponding coefficients for specific-binding (C-specific) values with Standardized Uptake Value Ratio (SUVR) values, we identified a significant correlation between the two. SUVR is a commonly used metric in PET imaging that represents the ratio of tracer uptake in a target region to a reference region, normalized over a specific time frame. Unlike SUVR, which requires a reference region, our method does not. This study underscores CIFA’s potential in advancing our understanding of tau dynamics, providing a robust and reference-region-independent alternative. Future research will aim to validate these findings with larger datasets and investigate their clinical relevance.