Tau PET
Tau PET imaging, using radiotracers to visualize tau protein aggregates in the brain, is crucial for diagnosing Alzheimer's disease but faces limitations due to cost and accessibility. Current research focuses on generating synthetic tau PET images from readily available MRI scans using various deep learning architectures, including generative adversarial networks (GANs) and diffusion models, often incorporating novel loss functions to improve image quality and accuracy. These advancements aim to increase the availability of tau PET data for research and potentially improve diagnostic capabilities, particularly by enabling MRI-free anomaly detection and analysis of tau protein spreading patterns through reaction-diffusion modeling.
Papers
October 28, 2024
June 18, 2024
May 21, 2024
July 16, 2023
June 21, 2023