Researchers say they can now use artificial intelligence to predict Alzheimer’s disease with nearly 93 percent accuracy.
Worcester Polytechnic Institute in Massachusetts announced Thursday that more than 800 brain scans helped AI identify anatomical changes in the brain that signal the onset of the most common forms of dementia.
The findings follow previous research showing that AI could help detect early risk factors for Alzheimer’s disease, predict risk, and identify undiagnosed Alzheimer’s patients.
Using AI to help diagnose Alzheimer’s disease early could give patients and doctors critical time to prepare and potentially slow the progression of the disease.
“Early diagnosis of Alzheimer’s disease can be difficult because symptoms can be mistaken for normal aging,” Benjamin Nephew, assistant professor at the institute, explained in a statement.
“However, we found that machine learning techniques can analyze large amounts of data from scans to identify subtle changes and accurately predict Alzheimer’s disease and related cognitive conditions,” he said.
MRI scans were collected from 344 people aged 69 to 84, of whom 281 had normal mental function, 332 had mild cognitive impairment, and 202 had Alzheimer’s disease.
The scans included 95 of about 200 different areas of the brain and used AI algorithms to predict the patient’s health status.
Analysis showed that one of the biggest predictors of Alzheimer’s disease is a decrease in brain volume. This occurs when brain cells in the hippocampus, which forms memories, the amygdala, which processes fear, and the entorhinal cortex, which gives us a sense of time, stop functioning.
This was true regardless of age and gender, with both men and women between the ages of 69 and 76 showing a decrease in brain volume in the right part of the hippocampus. This suggests that this region is important for early diagnosis, the researchers noted.
The study also showed that brain areas shrink differently depending on gender. In women, brain volume loss occurred in the left middle temporal cortex of the brain, which is involved in language and vision. In males, it was primarily found in the right entorhinal cortex.
Researchers believe this may be due to changes in sex hormones, such as decreased estrogen in women and testosterone in men.
Nephew said these conclusions could help improve future diagnostic and treatment methods. According to the Alzheimer’s Association, more than 7.2 million Americans are living with Alzheimer’s disease.
Further research is being conducted to uncover other influencing factors.
“A key challenge in this research is to build generalizable machine learning models that capture the differences between healthy brains and the brains of people with mild cognitive impairment or Alzheimer’s disease,” Nephew said.
