AI Algorithm Foresees Alzheimer’s Possibility Two Years Beforehand


Published Date : Aug 30, 2017

An artificial intelligence (AI)-driven algorithm has been discovered in a new study that could near-accurately predict individuals likely to develop Alzheimer’s disease in up to a two years of time in advance. It has also been anticipated to identify the early signs of dementia by reading brain scans. According to Sulantha Sanjeewa, a lead study author and McGill University computer scientist in Canada, the algorithm could particularly assist in picking patients for clinical trials conducted for drugs that are envisioned to delay the onset of Alzheimer’s disease. The algorithm has precisely predicted the probability of Alzheimer’s disease 84.0% of the time, according to Sanjeewa.

AI Program Selects Best Participants for Alzheimer’s Drug Trials

McGill University psychiatry, neurosurgery, and neurology associate professor, Dr. Pedro Rosa-Neto, also a co-lead author of the new study, has said that the technology could offer improved results than solely depending on humans, although it is still in a nascent stage. The findings of the study had been published online in the Neurobiology of Aging journal in July. Rosa-Neto also said that it could be possible for one to better test new medications for preventing Alzheimer’s disease if individuals in a group with a probable development of the disease have been identified.

Researchers have said that there could be three important steps in the creation of an effective AI algorithm, i.e. software writing, training, and testing. The engineers have designed the algorithm taking into account that only a fraction of a people could develop dementia. Furthermore, the programmers considered the probability of amyloid protein buildup occurring at different brain locations, in different concentrations, and at different rates.