Researchers at Worcester Polytechnic Institute (WPI) used a form of artificial intelligence (AI) to analyze anatomical changes in the brain and predict Alzheimer’s disease with nearly 93% accuracy. Their study, published in the journal Neuroscience, also found that the anatomical changes that accompany brain volume loss differed by age and gender.
“Early diagnosis of Alzheimer’s disease can be difficult because symptoms can be mistaken for normal aging,” says Benjamin Nephew, assistant professor in the Department of Biology and Biotechnology. “However, we now know that machine learning techniques can analyze large amounts of data from scans to identify subtle changes and accurately predict the cognitive conditions associated with Alzheimer’s disease. This advance will inform Alzheimer’s disease research and could lead to ways that doctors can diagnose and treat the disease earlier and more effectively.”
Alzheimer’s disease is a neurodegenerative disease that impairs mental function and ultimately leads to death. An estimated 6.9 million Americans over the age of 65 have Alzheimer’s disease.
A healthy brain contains billions of neurons, cells that process and transmit signals needed for thinking, movement, and other bodily functions. Alzheimer’s disease damages neurons, leading to cell death, loss of brain tissue and associated cognitive function.
Nephew doctoral student Sengbao Lu and Babin Jogeshwar (MS 24) conducted the study using brain MRI scans from the Alzheimer’s Disease Neuroimaging Initiative, a multicenter project to build a library of brain scans from ages 69 to 84. The scans depict the brains of people with normal mental function, mild cognitive impairment, and Alzheimer’s disease.
Analyzing data-rich MRI images can require significant computing power and time. To focus their investigation, WPI researchers first used machine learning to analyze 815 MRI scans for volumetric measurements in 95 brain regions. They then deployed an algorithm that makes predictions based on differences in measurements between healthy individuals and people with mild cognitive impairment or Alzheimer’s disease.
The results showed that the method could detect Alzheimer’s disease in normal brains and brains of people with mild cognitive impairment with an accuracy of 92.87%.
Volume loss in the hippocampus, amygdala, and entorhinal cortex was the greatest predictor of Alzheimer’s disease across age and gender categories. The hippocampus is a small, seahorse-shaped structure deep in the brain that is responsible for memory and learning. The amygdala is made up of two almond-shaped structures and controls emotions. The entorhinal cortex is a center for memory, navigation, and perception, and is one of the first parts of the brain to be affected by Alzheimer’s disease.
Men and women between the ages of 69 and 76, the youngest age group studied, showed decreased brain volume in the right hippocampus. Researchers say this suggests that the right hippocampus may be important for early diagnosis of Alzheimer’s disease.
“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. “A generalizable model means that the biomarkers we discovered are not unique to this dataset, but may be universal to all patients with mild cognitive impairment or Alzheimer’s disease.”
Differences between male and female brains have also been revealed. Researchers found that the women’s volume loss occurred in the left middle temporal cortex, which is involved in language, memory, and vision. In men, volume reduction in the right entorhinal cortex was significant.
The extent of these differences is surprising and may be related to an interaction between the progression of Alzheimer’s disease and changes in sex hormones, Nephew says. Some researchers have linked the risk of Alzheimer’s disease to age-related declines in estrogen in women and testosterone in men.
Nephew and his WPI students are following up on neuroscience publications by evaluating the use of deep learning models and investigating other factors that may influence the brain and Alzheimer’s disease (such as diabetes). The research brought together WPI students from a wide range of fields, from biology and biotechnology to neuroscience, psychology, computer science, and bioinformatics.
“This study demonstrates the strength of WPI neuroscience as being interdisciplinary and computational,” Nephew said. “The brain is an incredibly complex organ, and we need to think broadly about how to better understand, predict, and treat the diseases that afflict it.”
/Open to the public. This material from the original organization/author may be of a contemporary nature and has been edited for clarity, style, and length. Mirage.News does not take any institutional position or position, and all views, positions, and conclusions expressed herein are those of the authors alone. Read the full text here.
