A new British artificial intelligence system can predict with up to 92% accuracy which people with memory problems will eventually develop dementia within the next two years. This is another indication that artificial intelligence has enormous potential that makes it possible to diagnose various diseases at an early stage. The goal is not only to diagnose impending dementia early, but also to reduce the number of people who are misdiagnosed with dementia.
Researchers at the University of Exeter, led by Professor David Llewellyn, who published the study in the American medical journal JAMA Network Open, used data from 15,307 people with an average age of 72 with memory problems (of whom 1,568 had been diagnosed with Alzheimer’s disease or another form of Alzheimer’s disease). dementia over the next two years) to “train” a new machine learning algorithm to recognize the initial symptoms of dementia. The “intelligent” system has learned to detect clues hidden in the data, which the human eye, even a neurologist or other specialist, cannot recognize.
Read the scientific study in English here.
In addition, 130 diagnoses (8% of the total) were found to be incorrect, as they were subsequently reversed. Of these false positives for dementia, the algorithm was able to diagnose that 84% were unrelated to dementia. So the system can not only distinguish who may develop neurodegeneration of the brain in the future, but also improve the accuracy of the diagnosis, so that the undiagnosed person is not diagnosed as sick.
“We are now able to teach computers to accurately predict who will develop dementia over the next two years. We are also pleased to see that our machine learning method is able to identify patients who may have been misdiagnosed,” Llewellyn said.
The new system will be subject to further study to evaluate its practical usefulness and confirm its potential to improve clinical diagnosis and treatment of dementia.
source: ΑΠΕ-ΜΠΕ
Follow Skai.gr on Google News
And be the first to know all the news
