Alzheimer's Algorithm May Predict Who Gets the Disease

Alzheimer's disease has no cure, but getting a diagnosis as quickly as possible can allow patients to start symptom-delaying drugs. The problem with getting that diagnosis, though, is that the early stages of Alzheimer's can look a lot like mild cognitive impairment, which may or may not progress into Alzheimer's. A new study from McGill University may make it easier to distinguish what will develop into full-blown Alzheimer's. The researchers used data and artificial intelligence to predict whether someone with mild cognitive impairment would develop Alzheimer's, and were able to do so two years before the onset of dementia symptoms with 84 percent accuracy. The Alzheimer's algorithm takes data like memory test results, glucose metabolism in the brain, PET scans, cerebrospinal fluid and MRIs into account to predict whether patients with mild cognitive impairment will progress to Alzheimer's.