Collaborating with researchers at Texas’ UT Health Houston, Christopher Gaiteri, PhD will be helping to develop a deep-learning AI system to link brain imaging with cell-specific genetic factors to better understand the genetic architecture of Alzheimer’s disease and cognitive decline. This five-year, $6.4 million grant from the National Institute on Aging will bring together neuroimaging and genetic data that could help researchers learn how and why the disease develops. Chris is an Associate Professor of Psychiatry and Behavioral Sciences at Upstate, also serving as an Assistant Professor at Rush University in their Department of Neurological Sciences.

Christopher Gaiteri, PhD, is contributing to a project applying AI to different groups of data from Alzheimer's patients to try and uncover potential genetic links.

When asked to summarize the Actionable, Integrated and Multiscale (AIM-AI) project, Chris keys in on one word; “unfurl”.

“If you think about a flag unfurling, it spreads out and you can see more detail and how the different elements come together,” he explains. “That's how your DNA works. It's all tightly crammed together into chromosomes, but sometimes part of those chromosomes open; they become more accessible, and you can read off a copy of it. That ultimately turns into protein, and then that protein interacts with other proteins and builds your tissues.” By using AI to track this process, Chris and his collaborators hope to better understand the more complicated causes of sporadic Alzheimer’s disease.

This project could help shed light on how sporadic Alzheimer's disease (SAD) develops in patients. SAD is the most common form of Alzheimer's disease, accounting for more than 90% of cases. “With familial Alzheimer’s, you can go directly from DNA to Alzheimer's disease without having to think about how it unfurls through a sequence of steps. But for sporadic Alzheimer's, it involves the action of hundreds or thousands of different locations on DNA,” Chris says. “It could be helpful to try to understand these intermediate steps; hopefully it will help us understand how DNA shifts into disease.”

Chris's work at Rush University brought him to this project. Rush has one of the oldest Alzheimer’s research centers, which offers a unique data set.

“Because they’ve been studying Alzheimer’s and collecting samples for so long, we have neural imaging and brain samples from the same patients; having both of these types of data on the same person is like a Rosetta Stone.

“We examined these people where we measured both types of data and we found that there's a lot of synchronization,” Chris continues. “We can tell you to target this protein because it’s synchronized to that part of the brain. This is the powerful stuff that we've developed out of data from Rush.”

Another piece of the data; using MRI scans to study dynamic functional connectivity (dFC). dFC is a relatively new way of looking at brain activity; it can show the connections being made between different parts of the brain and how that changes in different brain “states”.

“In Alzheimer's disease, you lose executive control; you lose your ability to plan and carry out multiple steps. Planning and high-level control seems to be very much reflected by this reorganization of brain states, your ability to adapt to the world. We're going to look at these scans in conjunction with Alzheimer's patients, we're going to see how amyloid and tau measured in the brains of these people after they died relates to their ability to reconfigure their brain communication networks.”

Creating an AI framework to help combine all these different data sets is crucial, combining his collaborator’s DNA data with his imaging and molecular analysis. Being able to analyze this information together could have a massive impact on Alzheimer’s treatment.

“Maybe your ability to shift from one brain state to another brain state is affected in Alzheimer's disease,” Chris hypothesizes. “But we also know that the levels of several proteins are changing in the brains of these people. Potentially we could say ‘if you want to control that dynamic brain activity that's actually generating your cognitive life, you should control the levels of these proteins.’”

You can read more about this project here.

Learn more about Upstate’s Department of Psychiatry’s research here.