Sci-Fri AM: Imaging - 07: Symmetric geometric transfer matrix partial volume correction technique for emission...

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Processing

Background:

• Quantitative Positron Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) require partial volume correction (PVC) due to limited spatial resolution.
• Conventional PVC methods use region-based approaches with geometric transfer matrices (GTM) to account for spill over between segmented regions.

Purpose of the Study:

• To introduce and validate a novel analytically derived symmetric GTM (sGTM) method for partial volume correction.
• To compare the performance of the sGTM method against the conventional GTM method.

Main Methods:

• Developed an analytically derived symmetric GTM (sGTM) method focusing on spill over between regional spread functions (RSFs).
• Validated the sGTM method using a 3D digital brain phantom and a physical phantom with varying sphere sizes.
• Compared sGTM with GTM regarding accuracy, precision, noise propagation, and robustness to misregistration and point spread function (PSF) errors.

Main Results:

• The sGTM method demonstrated accuracy comparable to the GTM method, within 5% of true values.
• sGTM exhibited superior precision and noise propagation characteristics compared to GTM, particularly for structures smaller than 13 mm.
• The sGTM method proved more robust against misregistration and PSF estimation errors than the GTM method.

Conclusions:

• The analytically derived sGTM method provides an effective approach for partial volume correction in PET and SPECT.
• sGTM offers improved noise performance and robustness, making it a valuable advancement over traditional GTM methods, especially for small-volume quantification.