AIBullisharXiv – CS AI · 18h ago7/10
🧠
Reconstructing and forecasting disease trajectories of patients with Alzheimer's disease using routine data in resource-constrained settings
Researchers developed GNOVA, a machine learning framework combining GRU neural networks with Neural ODEs and variational autoencoders to predict Alzheimer's disease progression using only routine clinical data without expensive neuroimaging. The model successfully reconstructed patient cognitive trajectories and forecasted future cognitive states with high accuracy across 1,727 ADNI patients over 10 years, enabling deployment in resource-constrained healthcare settings.