A Unified Framework for Plasma Data Modeling in Dynamic Positron Emission Tomography Studies.

Area of Science:

• Nuclear Medicine
• Pharmacokinetics
• Biomedical Engineering

Background:

• Dynamic positron emission tomography (PET) requires accurate arterial plasma tracer concentrations for quantification.
• Current methods for plasma data modeling face challenges with radiometabolites, noise, and inconsistent, non-physiological models.
• Methodological variability in plasma modeling impacts the interpretation of PET kinetic studies.

Purpose of the Study:

• To develop and validate a unified, automated framework for arterial plasma data modeling in dynamic PET studies.
• To achieve accurate and robust description of plasma tracer kinetics.
• To reduce methodological variability in PET data analysis.

Main Methods:

• A novel pipeline using basis pursuit techniques for estimating radiometabolites and parent tracer concentrations.
• Testing the pipeline on four PET tracers ([11C]PBR28, [11C]MePPEP, [11C]WAY-100635, [11C]PIB) with both continuous and discrete blood sampling.
• Comparison with standard plasma data modeling procedures.

Main Results:

• The proposed pipeline demonstrated a similar fit for parent fraction but a superior description of total plasma radioactivity compared to standard methods.
• This improved plasma modeling led to a more accurate fit of the tissue PET data.
• The method showed superior fits for both continuous and discrete arterial sampling protocols.

Conclusions:

• The developed pipeline offers a robust, flexible, and user-friendly approach to arterial data modeling in dynamic PET.
• It has the potential to standardize blood data modeling across different studies and tracers.
• This standardization can enhance the reliability and interpretability of PET kinetic analyses.