Robust estimation of kinetic parameters in dynamic PET imaging

Area of Science:

• Nuclear Medicine
• Biomedical Engineering
• Pharmacokinetics

Background:

• Dynamic PET imaging is crucial for biological research, clinical diagnosis, and pharmacokinetic analysis.
• Kinetic parameters quantify radiotracer exchange and metabolism, but estimation is challenged by low count rates and arterial input function uncertainties.
• Quantitative analysis of individual kinetic parameters remains an unmet need.

Purpose of the Study:

• To develop a robust kinetic parameter estimation framework for dynamic PET.
• To address limitations posed by poor data statistics and arterial input function uncertainties.
• To enable quantitative analysis of individual kinetic parameters.

Main Methods:

• Utilized a robust H infinity estimation strategy optimized under the minimax criterion.
• Developed a framework to enhance robustness against measurement data properties and input function uncertainties.
• Validated the approach using Monte Carlo simulated data and real patient scans.

Main Results:

• The proposed framework demonstrates robustness to poor statistical properties in dynamic PET data.
• It effectively handles uncertainties in the estimated arterial input function.
• The method allows for quantitative analysis of individual kinetic parameters.

Conclusions:

• The developed framework significantly improves the accuracy and reliability of kinetic parameter estimation in dynamic PET.
• This advancement holds potential for enhanced biological research, clinical diagnosis, and pharmacokinetic analysis.
• The ability to quantitatively analyze individual parameters opens new avenues for personalized medicine and tracer development.