Posted on August 11, 2026 by Sean M. Wood
Klesse College of Engineering and Integrated Design biomedical engineer Dr. Amina Ann Qutub leads BEACON — Bedside Emergency AI for Care Orchestration project, an inaugural AI Moonshot Innovation Award from the UT REAL Health AI program.
UT San Antonio engineer leads UT researchers, clinicians and startup partners who are deploying AI tools to accelerate emergency care decisions, strengthen trauma systems and support recovery after brain injury and stroke.
Traumatic injuries are associated with over 40 million emergency department visits in the United States each year. In Texas, trauma is the leading cause of death among people under 46 and a major cause of long-term disability among survivors. A multidisciplinary team spanning artificial intelligence, biomedical engineering, trauma surgery and neurology is working to change that trajectory.
Led by Klesse biomedical engineer Dr. Amina Ann Qutub, with co-principal investigators and trauma surgeons Dr. Brian Eastridge and Dr. Elizabeth Scherer, and neurologist Dr. Mark Goldberg of UT Health San Antonio, the team received an inaugural AI Moonshot Innovation Award from the UT REAL Health AI program to launch BEACON — Bedside Emergency AI for Care Orchestration. Over the 18-month, $500,000 pilot project, BEACON will test, deploy and scale AI innovations designed to improve emergency care decision-making and outcomes for trauma and stroke patients across UT sites and their rural partners.
The project is part of a broader University of Texas investment in artificial intelligence across the UT health system and partner institutions. The UT System has committed $3.6 million to 12 projects across 11 campuses, including UT San Antonio.
Building AI for the trauma bay — with clinicians in the loop
During its first nine months, BEACON will test Selja, an AI clinical assistant developed by the team to support emergency triage and transfer decisions. Selja has been trained on more than 1 million trauma cases from across Texas, including prospective data from seven UT hospital system trauma centers. But the team says the system's development depends on far more than the scale of its data.
Clinicians are helping shape how the AI reasons.
“The clinical members of the team have been working closely with our AI expert colleagues to create a reliable and consistent decision tool for all providers caring for trauma patients,” said Scherer, Assistant Professor of Trauma and Emergency Surgery. “In environments with variable resources, the goal is to ensure patients get to the right level of care within the right amount of time.”
Qutub said that partnership between clinicians and engineers is fundamental to BEACON's approach. “What differentiates this from simply training on data is that we're incorporating feedback on the reasoning of our clinical colleagues,” said Qutub, the Burzik Professor in Engineering Design and associate professor in the Department of Biomedical and Chemical Engineering. “We’re training the AI to think like the doctors.”
The larger goal is to build a shared AI and data infrastructure that can improve decisions before a patient reaches the hospital, during acute care and throughout recovery.
The collaboration began in 2024 and was catalyzed by the Trauma Research and Combat Casualty Care Collaborative (TRC4). It brings together clinicians with deep experience in trauma, critical care and neurologic recovery alongside engineers and AI researchers working with an extensive body of trauma-related data. Scherer said the team has reviewed large numbers of patient cases and AI responses to ensure the system provides information that is clinically meaningful, consistent and actionable. “We have reviewed countless patient scenarios to ensure they are realistic and provide the information needed to support decision-making,” she said. “We have also evaluated numerous responses to make sure that, even when the wording differs, the content and clinical point remain consistent and actionable.”
Scherer, who serves as Surgical Critical Care Fellowship Director, is joined on the Selja clinical training team by Eastridge and Goldberg. Eastridge is Professor of Surgery at UT Health San Antonio, Division Chief of Trauma and Emergency Surgery, and the Jocelyn and Joe Straus Endowed Chair in Trauma Research. Goldberg is the LeWinn Endowed Professor in Neurology, and hespecializes in stroke and cerebrovascular disorders, with research focused on recovery after stroke and brain injury.
The collaboration deliberately brings together experts from very different environments.
“You have AI researchers and engineers who have never worked in a trauma bay working side by side with trauma surgeons, nurses and staff who have never worked with AI researchers,” Scherersaid. “The goal with BEACON is to augment the information available to clinicians when critical decisions have to be made.”
Minutes matter
In trauma care, delays can have lasting consequences.
Qutub said the team is focused on situations in which faster access to the right information could materially change a patient's trajectory.
“Every minute of delay in definitive trauma care can have serious consequences for a patient's recovery,” Qutub said. “If AI can help clinicians reach the right decision even one minute faster, that could translate into fewer disabilities and more lives saved.”
BEACON is not designed to replace physicians. Instead, the team is building AI to support highly trained clinicians with additional information, context and decision support. In essence, clinicians and biomedical / AI engineers are the architects building this new frontier by guiding emerging technologies to solve a significant public health problem.
Qutub compares the technology to an airplane's autopilot.
“You need the human pilot and copilot, even when the plane is flying on autopilot,” she said. “The pilots still must be extremely well trained. We're building an ecosystem that helps determinewhen a decision can be augmented by AI – and when it is better for the human to make that decision directly.”
Extending AI from emergency care to recovery
While the first phase of BEACON focuses on Selja and prehospital decision support, the second phase will extend the project beyond the emergency setting and into recovery.
The team will build a patient-centered data repository to follow recovery over time and develop AI advocates designed to support patients and caregivers after traumatic injury, particularly after brain injury and stroke.
The vision is a connected system in which the same infrastructure that helps clinicians make better decisions during an emergency can also help patients navigate rehabilitation, track progressand receive more personalized support after leaving the hospital.
To accomplish those goals within the 18-month pilot project period, BEACON is partnering with two university-affiliated startups.
PaloBio, a neuroresilience company, will develop personalized AI advocates for patients and their therapists, trainers, and caregivers to monitor and help in recovery. Selja AI, a UT spinout, will oversee Selja's performance and compatibility across software environments used throughout the UT health system.
Together, the partners aim to create a model in which AI strengthens every stage of the trauma journey from the first critical decisions about where a patient should receive care to the long process of rebuilding function and independence afterward.
For the BEACON team, the technology is not the endpoint. The goal is a trauma and stroke care ecosystem in which clinicians have better information, patients reach the right care faster, and survivors receive continuous support on the path to recovery.
— Sean M. Wood
