Approved research

Machine learning for Alzheimer's disease

Lay summary

We developed a quantitative method to visualize damaged brain tissues in patients experiencing Alzheimer's Disease and to predict the resultant aspects of cognitive decline, including memory, vision, reaction, and reasoning. The model can be employed to precisely diagnose each small subregion within the brain and distinguish brain regions with high risk of causing cognitive decline. Precise medical treatments can be made based on a detailed understanding of the degeneration of each patient's brain, introducing the possibility for individualized treatment of Alzheimer's and other neurodegenerative diseases. The outcome of our methods, notably the 3D risk map, can be broadly used for computer-assisted diagnosis as a critical visualization tool for fine spatial assessments of brain MRI.