PET AIF estimation when available ROI data is impacted by dispersive and/or background effects.

Area of Science:

• Nuclear Medicine
• Medical Imaging Physics
• Biomedical Engineering

Background:

• Positron Emission Tomography (PET) imaging relies on accurate arterial input function (AIF) extraction.
• Blood pool region of interest (ROI) data in PET can be compromised by dispersion and background effects.
• These artifacts hinder precise AIF signal quantification, impacting downstream kinetic modeling.

Purpose of the Study:

• To develop a novel, data-adaptive method for estimating the arterial input function (AIF).
• To address challenges in AIF extraction caused by dispersive and background effects in PET ROI data.
• To improve the accuracy and reliability of AIF estimation in PET studies.

Main Methods:

• Representing the AIF using the whole-body impulse response (IR) to the injection profile.
• Analyzing a population of directly sampled arterial data to characterize tracer IR statistical behavior.
• Developing a penalty term for regularized AIF estimation using quadratic programming.

Main Results:

• The impulse response (IR)-based method demonstrates computational efficiency.
• Evaluated with eight tracers in PET cancer imaging, including FDG, FLT, CO2, and H2O.
• Simulations show significantly reduced mean squared error compared to direct ROI methods, even with minor contamination.

Conclusions:

• The proposed IR-based AIF extraction scheme provides a practical solution for contaminated PET data.
• This method enhances the accuracy of AIF estimation in the presence of dispersion and background effects.
• It offers a robust approach for quantitative PET imaging, particularly in oncology.