FDG-NeuroSegmenter
FDG-NeuroSegmenter is a deep-learning-based model developed to perform the automatic segmentation of 52 anatomical regions in brain [18F]FDG PET images.
Here you can find the nnU-Net models! For more information and other resources check out the repository on GitHub.
If you use the FDG-NeuroSegmenter models in your research, please cite our paper:
Brain Fluorodeoxyglucose PET Anatomical Segmentation via AI: Extensive Validation in the Neurodegenerative Spectrum
Luísa C. Silva, Francisco P. M. Oliveira and Durval C. Costa for the Alzheimer's Disease Neuroimaging Initiative and for the Frontotemporal Lobar Degeneration Neuroimaging Initiative
Brain (2026)
DOI:10.1093/brain/awag314