Implementation of a Spatially-Variant and Tissue-Dependent Positron Range Correction for PET/CT Imaging

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Reconstruction

Background:

• Positron range (PR) in PET imaging introduces blurring, affecting image quality.
• Accurate correction for PR is crucial, especially for higher energy positron emitters.
• Existing methods may not adequately account for tissue-dependent PR variations.

Purpose of the Study:

• To develop and evaluate a novel, spatially variant, and tissue-dependent positron range correction (PRC) method.
• To improve image reconstruction in PET by addressing PR blurring.
• To enhance the diagnostic accuracy of PET imaging.

Main Methods:

• Simulated PR distributions for 18F, 68Ga, and 124I in various materials (lung, water, bone) using GATE.
• Developed spatially variant kernels from uniform kernels based on surrounding material composition.
• Implemented tissue-dependent point-spread functions into iterative image reconstruction (OSEM).
• Evaluated the PRC method using NEMA image quality and dedicated PR phantoms.

Main Results:

• PRC had a negligible effect on 18F imaging.
• Image contrast improved by 33% (68Ga) and 24% (124I) for a 10-mm sphere.
• Contrast recovery improved by 5% for larger spheres.
• Spatial resolution improved by 26% for 124I (FWHM 4.9 to 3.7 mm).

Conclusions:

• The proposed spatially variant, tissue-dependent PRC method effectively improves PET image quality.
• This approach enhances contrast and spatial resolution, particularly for high-energy positron emitters.
• The method holds potential for improving clinical and research PET data quality.