Houston researchers create AI model to tap into how brain activity relates to illness
brainiac
Houston researchers are part of a team that has created an AI model intended to understand how brain activity relates to behavior and illness.
Scientists from Baylor College of Medicine worked with peers from Yale University, University of Southern California and Idaho State University to make Brain Language Model, or BrainLM. Their research was published as a conference paper at ICLR 2024, a meeting of some of deep learning’s greatest minds.
“For a long time we’ve known that brain activity is related to a person’s behavior and to a lot of illnesses like seizures or Parkinson’s,” Dr. Chadi Abdallah, associate professor in the Menninger Department of Psychiatry and Behavioral Sciences at Baylor and co-corresponding author of the paper, says in a press release. “Functional brain imaging or functional MRIs allow us to look at brain activity throughout the brain, but we previously couldn’t fully capture the dynamic of these activities in time and space using traditional data analytical tools.
"More recently, people started using machine learning to capture the brain complexity and how it relates it to specific illnesses, but that turned out to require enrolling and fully examining thousands of patients with a particular behavior or illness, a very expensive process,” Abdallah continues.
Using 80,000 brain scans, the team was able to train their model to figure out how brain activities related to one another. Over time, this created the BrainLM brain activity foundational model. BrainLM is now well-trained enough to use to fine-tune a specific task and to ask questions in other studies.
Abdallah said that using BrainLM will cut costs significantly for scientists developing treatments for brain disorders. In clinical trials, it can cost “hundreds of millions of dollars,” he said, to enroll numerous patients and treat them over a significant time period. By using BrainLM, researchers can enroll half the subjects because the AI can select the individuals most likely to benefit.
The team found that BrainLM performed successfully in many different samples. That included predicting depression, anxiety and PTSD severity better than other machine learning tools that do not use generative AI.
“We found that BrainLM is performing very well. It is predicting brain activity in a new sample that was hidden from it during the training as well as doing well with data from new scanners and new population,” Abdallah says. “These impressive results were achieved with scans from 40,000 subjects. We are now working on considerably increasing the training dataset. The stronger the model we can build, the more we can do to assist with patient care, such as developing new treatment for mental illnesses or guiding neurosurgery for seizures or DBS.”
For those suffering from neurological and mental health disorders, BrainLM could be a key to unlocking treatments that will make a life-changing difference.
Anshumali Shrivastava is an associate professor of computer science at Rice University. Photo via rice.edu