Symmetric geometric transfer matrix partial volume correction for PET imaging: principle, validation and robustness

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Processing

Background:

• Positron emission tomography (PET) imaging is limited by spatial resolution, necessitating partial volume correction (PVC) for accurate quantification.
• Conventional region-based PVC methods utilize anatomical images and geometric transfer matrices (GTM) to correct for spill-over between regions.

Purpose of the Study:

• To introduce and validate a novel, analytically derived symmetric GTM (sGTM) method for PET partial volume correction.
• To compare the performance of the sGTM method against the conventional GTM method in terms of accuracy, precision, noise propagation, and robustness.

Main Methods:

• Developed an analytically derived symmetric GTM (sGTM) method based on spill-over between regional spread functions.
• Validated the sGTM method using digital sphere and brain phantoms, and a physical sphere phantom.
• Assessed performance by introducing mis-registration and point spread function (PSF) estimation errors.

Main Results:

• The sGTM method demonstrated accuracy comparable to the GTM method (within 5%) across all phantoms.
• sGTM exhibited superior precision and noise propagation compared to GTM, particularly for structures smaller than 13 mm.
• sGTM showed improved robustness against mis-registration and PSF errors, especially for smaller objects.

Conclusions:

• The analytically derived sGTM method provides accurate PET quantification.
• The sGTM method offers enhanced precision, reduced noise propagation, and superior robustness compared to conventional GTM methods.
• sGTM is a promising advancement for quantitative PET imaging, particularly for small lesion detection and characterization.