Assessment of S values in stylized and voxel-based rat models for positron-emitting radionuclides.

PURPOSE

Positron emission tomography (PET) is a powerful tool in small animal research, enabling noninvasive quantitative imaging of biochemical processes in living subjects. However, the dosimetric characteristics of small animal PET imaging are usually overlooked, although the radiation dose may be significant. The variations of anatomical characteristics between the various computational models may result in differences in the dosimetric outcome.

METHODS

We used five different anatomical rat models (two stylized and three voxel based) to compare calculated absorbed fractions and S values for eight positron-emitting radionuclides (C-11, N-13, O-15, F-18, Cu-64, Ga-68, Y-86, and I-124) commonly used to label various probes for small animal PET imaging. The MCNPX radiation transport code was used for radiation dose calculations.

RESULTS

For most source/target organ pairs, O-15 and Ga-68 produce the highest self-absorbed S values because of the high-energy and high-frequency of positron emissions, while Y-86 produces the highest cross-absorbed S values because of the high energy and high frequency of γ-rays emission. Anatomical models produced from different rat strains or modeling techniques exhibit different organ masses, volumes, and thus give rise to different S values and absorbed dose. The variations of absorbed fractions between models of the same type are less than those between models with different types. The calculated S values depend strongly on organ mass, and as such, different models produce similar S values for organs of comparable masses. In most source organs presenting with high cumulated activity, the absorbed dose is less affected by model difference compared with other organs.

CONCLUSIONS

The produced S values for common positron-emitting radionuclides can be exploited in the assessment of radiation dose to rats from different radiotracers used in small animal PET experiments. This work contributes to a better understanding of the influence of different computational models on small animal dosimetry.