Partial-volume effect correction in positron emission tomography brain scan image using super-resolution image...

Area of Science:

• Medical Imaging
• Image Reconstruction
• Quantitative Analysis

Background:

• Partial-volume effect (PVE) in positron emission tomography (PET) arises from limited spatial resolution, causing underestimation of activity concentrations.
• This underestimation is particularly significant in brain imaging when source dimensions are small relative to the system's resolution.
• PVE can lead to misinterpretation of PET scan results.

Purpose of the Study:

• To characterize the effectiveness of super-resolution (SR) image reconstruction in compensating for PVE in PET.
• To evaluate the ability of SR to improve quantitative accuracy in small structures within PET images.

Main Methods:

• Evaluated SR image reconstruction's ability to recover activity concentrations in small structures.
• Compared images before and after SR reconstruction using NEMA/IEC and Hoffman brain phantoms.
• Assessed performance in four human brain subjects (three normal, one atrophic) using apparent recovery coefficient (ARC) and percentage yield.

Main Results:

• SR implementation significantly improved both ARC and percentage yield in phantom studies.
• SR enhanced image intensity and demonstrated efficiency in correcting PVE in normal subjects.
• The study confirmed SR's capability to improve quantitative accuracy in brain PET.

Conclusions:

• Super-resolution (SR) algorithms can effectively correct partial-volume effect (PVE) in PET images.
• SR enhances quantitative accuracy, particularly for small structures in brain PET.
• Future research will expand SR application to whole-body PET imaging.