Standardization and quality control of brain PET data in a multicenter study

Abstract

PET data depends on the PET camera model and the detailed methods of PET scanning including subject preparation, injected activity, accumulation time, scan time, and reconstruction parameters. In a multicenter study, data quality tends to be heterogeneous because participating PET centers have various PET camera models with variable experience in brain PET. Therefore standardization of the PET imaging protocol is essential, and the physical image quality should be harmonized between different camera models based on phantom tests. Japanese Society of Nuclear Medicine (JSNM) has issued the standard PET imaging protocols and determined phantom test criteria to standardize and harmonize brain FDG-PET and amyloid imaging. JSNM has further launched a site qualification program that qualifies PET centers and their cameras for their capability of high-quality standardized PET imaging based on phantom data and site visit. This program can be used to select or qualify PET centers and cameras in a multicenter project before initiation of the study. In such a multicenter study, each PET data should be checked for protocol compliance and data quality by PET QC Core before being forwarded to data interpretation and analysis.