Quantitative analysis of template-based attenuation compensation in 3D brain PET

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiology

Background:

• 3D brain PET imaging often requires attenuation correction (AC) for accurate quantification.
• Traditional AC uses patient-specific transmission scans, which are time-consuming.
• A novel atlas-guided AC method eliminates the need for transmission scans.

Purpose of the Study:

• To assess the quantitative accuracy of the atlas-guided AC method in a clinical setting.
• To evaluate the performance of atlas-guided AC for individual quantification in brain PET.
• To compare atlas-guided AC with the gold standard transmission-based AC.

Main Methods:

• The study utilized automated volume of interest (VOI)-based analysis with BRASS software.
• Clinical brain PET data underwent both atlas-guided and measured transmission-guided AC.
• Quantitative accuracy was assessed by comparing results from both AC methods.

Main Results:

• A strong correlation (R²=0.91) was observed between atlas-guided and measured transmission-guided AC.
• The regression line showed good agreement with the line of identity (slope=0.96).
• Mean relative difference was 2.3%, with a maximum difference <7%, and no statistically significant regional differences were found.

Conclusions:

• The atlas-guided AC method demonstrates high quantitative accuracy for individual brain PET analysis.
• This method is a reliable alternative to transmission scans for research and diagnostic applications.
• The findings support the use of atlas-guided AC for improved efficiency and accuracy in brain PET studies.