Your words can predict your brain’s future | Adolfo García | TEDxRiodelaPlata GIS

García opens with a powerful story about Susie, a woman with Alzheimer's who can't recognize her own son, illustrating the devastating impact of this disease that affects 40 million people worldwide. He emphasizes the urgency of the Alzheimer's crisis, noting that cases will triple by 2050, with odds reaching one in three by age 85, yet no cure exists. The speaker identifies early detection as our best weapon against Alzheimer's, enabling timely intervention, but criticizes current diagnostic methods as lengthy, stressful, costly, and inaccessible to most people globally. This problem is particularly acute in low and middle-income countries, which account for 60% of cases but receive less than 25% of research funding. García then introduces his solution: AI-powered speech analysis that can detect Alzheimer's traces in natural conversation. His research team, comprising neurolinguists, clinicians, and data scientists, has trained machine learning models to identify speech patterns characteristic of Alzheimer's patients. Their breakthrough achievement includes 90% accuracy in detecting Alzheimer's from one-minute speech recordings, discovering that patients tend to use simpler, more frequent words with common sound structures. The technology has been successfully validated in various settings including homes, hospitals, and phone calls, with some clinicians already using it to inform diagnostic decisions. García envisions a democratized future where this technology provides equitable early detection access regardless of economic status or location, calling for global dialogue among clinicians, scientists, and policymakers to make these innovations universally available.