MIT Aging Brain Symposium Details AI Alzheimer's Care and Anesthesia Work
At MIT's Aging Brain Initiative symposium, researchers described an AI model for personalized Alzheimer's care, anesthesia dosing research, and glycan studies.
The story in focus
MIT's Aging Brain Initiative held its Innovations for the Aging Brain symposium on Sept. 25, where six talks covered projects including an AI system to match Alzheimer's patients to prevention interventions, brain-wave-guided anesthesia dosing, and glycan research tied to Alzheimer's.
- An AI model called FINGERS-7B integrates clinical trial, lifestyle and multi-omics data to predict Alzheimer's progression and likely response to interventions.
- Anesthesia research has shown multiple drugs induce unconsciousness by shifting brainwaves out of phase, informing a closed-loop dosing system.
- A lab has shown glycans of different sequences organize on cell membranes to bind and recruit specific proteins.
AI for personalized Alzheimer's prevention
Giovanni Traverso, a professor of mechanical engineering and physician at Brigham and Women's Hospital, has worked with collaborators including ABI Director Li-Huei Tsai on an AI system intended to improve personalized Alzheimer's care by matching patients to the best prevention interventions for them.
The system integrates clinical trial data, patient lifestyle histories and multi-omics data such as genomics and proteomics to predict an individual's Alzheimer's progression and likely response to interventions. Speaking alongside project leads, Traverso described the model, called FINGERS-7B, which was trained on 8 trillion tokens drawn from those diverse sources. Collaborator Adrián Noriega de la Colina noted that individual biology is unique and links to how people respond to interventions, and said predictions could also inform clinical trial designs.
Anesthesia, lab models and molecule probes
Earl K. Miller described work with Emery N. Brown on a closed-loop system that uses brain wave-based measurements of unconsciousness to optimize anesthesia dosing, citing high rates of postoperative delirium in older adults after major surgery. Their labs have shown that multiple anesthesia drugs induce unconsciousness by shifting brainwaves out of phase.
Ritu Raman discussed lab models of muscle tissue that integrate nerves and vasculature, made for ALS from patient-derived induced pluripotent stem cells, to examine neuromuscular junctions and drug responses. Oleta Johnson focuses on preventing protein aggregation before it starts, while Alan Jasanoff develops chemical probes to make molecules in the brain visible in MRI.
Glycans and a first observation
Laura Kiessling discussed glycans, sugars lining cell surfaces, and the link researchers have drawn between heparan sulfate and Alzheimer's disease. Her lab, working with Ed Boyden, has shown that glycans of different sequences organize on cell membranes so they can bind and recruit specific proteins to distinct cell regions. Kiessling said the team was excited about the first observation and about having a toolkit to study other diseases.
What is the FINGERS-7B model designed to do?
It is an AI model built to predict an individual's Alzheimer's progression and likely response to prevention interventions by integrating clinical trial, lifestyle and multi-omics data, according to symposium researchers.
Sources & further reading
Help keep the record accurate. Send a correction.
