Evaluation of applying space-variant resolution modeling to attenuation correction in PET

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Reconstruction

Background:

• Attenuation correction is crucial in positron emission tomography (PET) for accurate quantification and image quality.
• Traditional methods using ray-tracing projection of attenuation maps are limited by unaddressed physical effects like finite crystal size and inter-crystal scatter.
• These limitations lead to underestimated tracer uptake and reduced image contrast.

Purpose of the Study:

• To evaluate the impact of incorporating space-variant resolution modeling (RM) into the attenuation correction process for PET.
• To address the physical effects typically omitted in conventional projection models for attenuation correction.
• To improve the accuracy of tracer uptake quantification and image contrast recovery in PET.

Main Methods:

• Implementation of space-variant RM to complement physical effects in attenuation correction.
• Verification of the developed method using Monte Carlo simulations on a custom-built brain PET scanner.
• Validation with real-world data from the brain PET scanner.
• Comparison against established methods: space-invariant Gaussian RM, average-depth-of-interaction, and multi-ray tracing.

Main Results:

• Space-variant RM demonstrated superior performance in reducing image artifacts.
• The method significantly enhanced image contrast recovery compared to traditional approaches.
• Results were consistent across both simulated and real-world PET data.
• Outperformed space-invariant Gaussian RM, average-depth-of-interaction, and multi-ray tracing methods.

Conclusions:

• Space-variant resolution modeling is an effective enhancement for PET attenuation correction.
• This advanced technique overcomes limitations of traditional methods by accounting for omitted physical effects.
• The improved accuracy in attenuation correction leads to better image quality, reduced artifacts, and more reliable quantification in PET imaging.