Modelling arterial input functions in positron emission tomography dynamic studies.

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Pharmacokinetics

Background:

• Dynamic positron emission tomography (PET) imaging requires accurate arterial input function (AIF) quantification for reliable tracer concentration analysis.
• Current methods often rely on invasive arterial blood sampling or mathematical models (e.g., tri-exponential, Feng's) that may not fully capture AIF complexity or injection dynamics.

Purpose of the Study:

• To develop and evaluate novel mathematical models for AIF quantification that explicitly incorporate radiotracer injection duration.
• To compare the performance of these new models against established methods using diverse PET datasets.

Main Methods:

• Proposed two new mathematical models for AIF modeling, integrating injection duration as a parameter.
• Validated model performance across eight distinct datasets from various PET facilities.

Main Results:

• The proposed models demonstrated improved ability to describe complex AIF behaviors, including varying tracer clearance rates.
• Incorporating injection duration enhanced the accuracy and robustness of AIF estimation compared to conventional models.

Conclusions:

• The novel AIF models offer a more accurate and comprehensive approach to dynamic PET data analysis.
• These models have the potential to improve the quantification of tracer kinetics and reduce reliance on invasive measurements.