Meta is committed to fostering a future that is inclusive for everyone, including the universities, professors, students, and communities in America that benefit from academic research. To support this vision, Meta is introducing Superintelligence models, research funding, and collaborative opportunities with researchers to the academic sphere by partnering with five US universities: California Institute of Technology, The Ohio State University College of Medicine, Louisiana State University, Indiana University, and UTHealth Houston.
The company is enhancing its academic partnerships by allocating $2 million across these institutions to facilitate research utilizing Meta’s models to propel advancements in fields such as medicine, physics, chemistry, biology, computer science, and mathematics.
In the realm of medicine and healthcare, researchers are tackling a variety of projects utilizing Meta’s models, glasses, and mixed reality headsets to address current challenges. Here are some examples:
Monitoring for Infant Tracheostomies
Infant tracheostomy emergencies arise when tracheostomy tubes become dislodged, posing life-threatening risks, especially since affected babies are unable to cry for help. Continuous monitoring by trained staff is often not feasible. Dr. Michael Dunham from Louisiana State University is spearheading the Eyes-On project, which aims to develop a computer vision system that continuously detects tracheostomy tube displacement and signs of distress in infants who cannot vocalize reliably. This system is experimenting with Meta’s open weight SAM models for tracking and is being designed for integration with bedside cameras and edge computing. A proof-of-concept has already been published, and further developments are underway.
Conversational AI for Dementia Care
Individuals with dementia frequently encounter communication issues that conventional conversational AI systems fail to manage effectively, causing frustration rather than assistance. Dr. David Crandall at Indiana University is working on a project to develop multimodal AI methods for analyzing conversations involving older adults with dementia. The goal is to enable conversational agents, including social robots, to tailor their interactions based on individual user needs. This research examines whether AI systems can be sensitized to the unique communication patterns of people with cognitive differences.
Identifying Heart Failure Risk Earlier
Heart failure with preserved ejection fraction (HFpEF) constitutes about half of all heart failure cases, yet recognizing it early is challenging due to symptom overlap with obesity and other conditions. As a result, formal evaluation often only begins after clinical suspicion arises. At The Ohio State University College of Medicine, AI researcher Dr. Yuan Xue, cardiologist Dr. Yuchi Han, and AI researcher Dr. Xia Ning are leading a project to explore how Meta’s open-weight models can anticipate HFpEF risk by analyzing routine abdominal CT scans, longitudinal medical records, and ECG data when available. The team is creating a model that accommodates missing data, a common occurrence in real-world clinical settings.
Further information on these projects can be found on our Research Blog, and we are eager to share more developments as these projects progress.