Covariance statistics and network analysis of brain PET imaging studies
The analysis of brain structural and functional imaging data using graph theory has increasingly become a popular approach since brain topological representations have proven to be helpful for visualising and understanding anatomical and functional relationships between different cerebral areas.
NetPET is a tool for graph-based analysis of brain PET studies which exploits population-based covariance matrices. NetPET aims to explore topological tracer kinetics differences in cross-sectional investigations. Simulations, test-retest studies and applications to cross-sectional datasets from three different tracers ([18F]FDG, [18F]FDOPA and [11C]SB217045) and more than 400 PET scans have been used to test the applicability of NetPET in healthy controls and patients. Results showed good reproducibility and general applicability of the methods within the range of experimental settings typical of PET neuroimaging studies.
- PET covariance analysis can be an alternative way to look into neuroimaging studies (complementary to standard cross-sectional analysis)
- Be aware of limitations (e.g. group homogeneity or sensitivity experimental variables)
- Plurality of alternatives (dynamic vs statistic method, generation of adjacency matrix)
NetPET is a Matlab-based software package, that can be freely downloaded through NITRC portal.
Reference: Veronese, M., Moro, L., Arcolin, M. et al. Covariance statistics and network analysis of brain PET imaging studies. Sci Rep 9, 2496 (2019). https://doi.org/10.1038/s41598-019-39005-8