Synthesizing PET images with ASL CBF for AD diagnosis

Motivation: ASL has the potential to replace PET in the diagnosis of AD. Goal(s): Developing a diffusion model to generate PET images from ASL CBF for both normal subjects and AD patients. Approach: Latent diffusion models were applied to use ASL CBF as condition to generate FDG and Amyloid PET. Results: The generated images show reasonable fidelity, across brain regions relevant for the diagnosis of AD. Impact: This study provides the proof-of-concept that noninvasive ASL can be applied to synthesize PET images for the diagnosis and management of AD without the cost and radiation exposure of an actual PET scan.