<< Back

What Happens When Big Data Is Applied to Diagnosing, Treating Mental Health Issues?

April 27, 2021

Big data sounds daunting – millions of factoids taking months to crunch through massive databases — but when Gregory Book extracts interesting statistics, the effort proves intriguing.

When he crafts charts showing how those tidbits factor into everything from the development of medication to an understanding of weather’s impact on the brain, the hook for neuroscientists grows stronger.

Book, a researcher with the Olin Neuropsychiatry Research Center at the Institute of Living, offered insight at a recent Grand Rounds conference on the impact of big data on progress in the diagnosis and treatment of mental health conditions.

“Clinically, big data has very real impact on what we do. I call it ‘an answer in search of a question,’” Book said. “Some signals are only visible with massive amounts of data.”

There’s a rich history of big data’s contributions to the evolution of medicine and science, he said, noting the National Institute of Mental Health Data Archive (NDA) began as the National Database for Autism Research. Since its metamorphosis, it has become a data archive with 1,728 independent projects sharing data.

At Olin, the data collection began with the first MRI in 2003. Since then, Book said affiliated scientists collected data for about 150 funded research projects using such equipment as the fMRI (functional magnetic resonance imaging, which measures brain activity), diffusion imaging and spectroscopy.

“Four gigabytes a day of data has grown to 15 gigabytes a day that is collected and stored,” he said. “That equals about 30 terabytes of data we’ve collected now.”

To demonstrate the information that can be gleaned from such massive stores of data, he pointed to a project on the impact of weather on brain size and function. Comparing hourly reports from the National Weather Service at Bradley International Airport with Olin brain scans conducted at the same time have prompted interesting deductions, he said.

“Previous research found that brain volume changes based on time of day in multiple sclerosis and Alzheimer’s patients,” he said. “Their brains are larger in the morning. Other studies also noticed seasonal changes — with maximum light in the summer, people do better on cognitive tests.”

Until that point, there had been no studies focused on the effect of weather on the brain.

“We formed a hypothesis, big data-style, and looked for studies with similar context,” Book said, noting the paper published that compares weather between 2004 and 2019 to 3,300 imaging studies during the same time.

“We found the barometric pressure affects the brain size,” he said. “The volume of the cerebellum goes down as the temperature goes up.”

More research is needed to fully understand the effect, a typical progression when mining big data, according to Dr. Godfrey Pearlson, director of Olin.

“It’s exciting the possibilities that accrue from having massive amounts of data,” he said.