DENVER (KDVR) — There’s new research showing you might be able to predict whether someone is susceptible to developing Alzheimer’s disease later on in life based off of their writing style.
The fascinating discovery was uncovered by researchers at IBM who used an artificial intelligence program to sift through writing tests submitted by 80 men and women who were in their eighties.
“It adds to a lot of research and work that has already been done in this space,” said Dr. Peter S. Pressman, a neurologist at UCHealth.
By using certain algorithms the researchers were able to pinpoint a specific portion of the group who featured similarities in their writing styles.
“Certain words that we use everyday in speech, like: ‘a’, ‘and’ – short little words began to fade away, drop out, as speech became more simple over time in writing,” Dr. Pressman.
The spelling and grammatical errors were identified in nearly every subject who ended up developing Alzheimer’s.
In Colorado alone, there are roughly 76,000 adults over the age of 65 who have Alzheimer’s. That figure is expected to climb to 92,000 by 2025, according to Alzheimer’s Association of Colorado.
“I think this is really important because we do really need to diagnosis Alzheimer’s as early as possible,” Dr. Pressman said.
According to the data collected, the A.I. program’s prediction had a 75% accuracy rate.
“75% accuracy is pretty good for a machine learning algorithm. It would not be considered good enough I think to integrate into everyday care right now. So there is more work that is going to be necessary,” Dr. Pressman said.
To read more about the new study, click here.