The IBM artificial intelligence algorithm is able to help diagnose Alzheimer’s disease early, with 75% accuracy, years before the onset of symptoms. The new algorithm can predict a person’s future with Alzheimer’s disease by examining a person’s handwriting and notes and some components.
This is the first time IBM researchers have announced that their new algorithm can predict the future of healthy, completely normal people years before the onset of symptoms.
According to the report «New York Times“This new approach could help doctors as a starting point in the early diagnosis of Alzheimer’s so that, if necessary, the necessary arrangements can be made for such people to avoid more serious problems.”
The study involved 270 volunteers and looked at their status over several years. Behavioral patterns of this group such as speech status, typing errors, forgetfulness and some other cases were recorded that can be cited as symptoms of nerve cell destruction disease.
In another experiment, researchers looked at a group of 80 men and women in their 80s, half of whom had Alzheimer’s, all of whom were in perfect health about 7.5 years ago and had no problems. Of these volunteers, the other half were still in perfect health and had no problems.
The team also participated in the Framingham Heart Study to monitor cardiovascular disease. Candidates were asked to take a writing test and write on a variety of topics, such as a description of a boy standing on an unstable wooden board or a child trying to pick up a dish from the kitchen cabinet and not reach his or her height.
Using artificial intelligence algorithms, the researchers sought to understand the subtle differences in their ability to use their mother tongue and to find out what their vocabulary is and how they use it. It was important for the researchers to know whether these people used words repetitively in writing the texts and what the participants’ typographical errors were like.
The results showed that among the participants, a group of them had various problems in writing the text, for example, they had many typos, repeated the same words many times in the text and did not have enough mastery of grammar and simple mistakes were seen in their texts.
IBM researchers have found that these people are more likely to develop Alzheimer’s in the future. As mentioned, the accuracy of this algorithm based on artificial intelligence is 75%.
Dr. Michael Wiener, an Alzheimer’s disease specialist at the University of California, Berkeley, says IBM is testing other components for early detection of Alzheimer’s and that the issue will not end there.