Experimental determination of PET modulation transfer functions incorporating positron range effects using an oversampled cylindrical source.

BACKGROUND

Spatial resolution is a fundamental performance parameter of positron emission tomography (PET) systems and is commonly assessed using standardized point-source-in-air procedures. While this configuration minimizes scattering and attenuation effects, it does not account for positron transport in tissue-equivalent media, whose contribution may become increasingly relevant as PET system resolution improves.

PURPOSE: To develop and validate a direct experimental methodology for PET spatial resolution assessment that incorporates positron-range effects without relying on Monte Carlo simulations or parametric modeling of the system response.

METHODS

A modular polymethyl methacrylate (PMMA) phantom containing a thin linear radioactive source ( 18 F or 68 Ga ) was used to ensure full positron absorption. By slightly inclining the source relative to the scanner axis, an oversampling strategy was implemented. The reconstructed volume was projected onto a one-dimensional profile to obtain a sub-pixel sampled point spread function (PSF), from which the modulation transfer function (MTF) was derived by Fourier transformation. Finite source-size correction and uncertainty propagation from PSF to MTF were performed. Radial and tangential MTFs were also determined at off-center positions.

RESULTS

The oversampling strategy yielded very small statistical uncertainties in both PSF and MTF estimates. Measurements performed on two PET/CT systems of different technological generations showed negligible differences between 18 F and 68 Ga in the older scanner, whereas measurable radionuclide-dependent effects were observed in the newer system. In the latter, 68 Ga exhibited a broader PSF and lower MTF than 18 F , with MTF 50 values of 0.085 and 0.112 mm - 1 , respectively.

CONCLUSIONS

The proposed methodology provides a statistically efficient framework for experimental PET spatial-resolution characterization under realistic conditions. It enables quantitative assessment of positron-range effects and complements standardized resolution measurements, particularly in modern high-resolution PET systems.