#IBM and #Pfizer researchers have developed an #artificialintelligence system that can predict the onset of #Alzheimer ‘s disease by analyzing voice samples. The system uses language processing algorithms to detect grammatical and syntactic errors in speech, which indicate incipient #dementia. It was tested on voice samples from the long-running US Framingham study and was able to predict the onset of the disease with 70% accuracy, even up to 7 years before clinical diagnosis. Early diagnosis is vital for Alzheimer’s patients as it enables them to access appropriate treatment and support, which is more effective when started sooner.
Fiona Carragher, research director of the Alzheimer’s Society, praised the research as an important step in the use of AI for accurate and timely diagnoses but stressed the need for further testing on larger and more diverse groups. More research is desperately needed as Alzheimer’s causes brain changes up to 15 years before symptoms appear.
source