Extraction of input function from rat [18F]FDG PET images.

Area of Science:

• Nuclear medicine
• Preclinical imaging
• Biomedical engineering

Background:

• Small animal positron emission tomography (PET) using 2-deoxy-2-[18F]fluoro-D-glucose ([18F]FDG) visualizes glucose uptake in rodents.
• Quantifying glucose uptake requires an input function, typically derived from arterial plasma radioactivity measurements.
• Extensive blood sampling can alter the physiological processes under investigation.

Purpose of the Study:

• To develop and validate a novel model-based technique (K-Model) for estimating the input function in rat [18F]FDG PET studies.
• To compare the K-Model's performance against the established F-Model and traditional blood sampling methods.

Main Methods:

• Two groups of rats with differing physiological conditions were scanned using HRRT and Inveon-PET/CT scanners.
• Input functions were estimated using both the F-Model and K-Model via an optimization procedure with boundary condition constraints.
• The glucose influx rate (Ki) was calculated using both estimated and measured input functions for validation.

Main Results:

• Both models accurately reproduced input functions using single-point blood count data.
• The K-Model demonstrated the highest feasibility, with a minimal difference of 1.1±15.1% in Ki values compared to blood sampling methods.
• Regression analysis revealed a strong correlation between image-based and blood sampling methods, with a slope near unity and intercept near zero.

Conclusions:

• The K-Model effectively estimates the input function from rat [18F]FDG PET images.
• This technique allows for glucose metabolism assessment without the physiological disturbances caused by frequent blood sampling.
• The K-Model offers a non-invasive approach to studying glucose metabolism in preclinical models.