Ai predicts Alzheimer's with 92% accuracy!

Researchers at the University of Toronto’s Tanz Centre have developed an AI tool that can detect Alzheimer’s up to a decade before symptoms appear. Using a convolutional neural network (CNN) trained on thousands of MRIs, the system identifies subtle brain changes invisible to human eyes—tiny texture and connectivity shifts that signal early neurodegeneration. It predicts with 92% accuracy who will develop Alzheimer’s, using routine, low-cost MRI scans instead of expensive PET tests. This breakthrough enables earlier intervention, more effective drug trials, and massive healthcare savings by leveraging existing infrastructure. Now undergoing broader validation and Health Canada approval, this innovation places Toronto at the forefront of AI-driven medical discovery and the fight against neurodegenerative disease.