View Larger Image
Joseph Sanford, M.D., and Adria Abella Villafranca use code they have written to test the OAK-D camera, equipped with spatial artificial intelligence.
Image by Karmen Robinson
Institute for Digital Health & Innovation Team Chosen as a Finalist in Artificial Intelligence Competition
| The University of Arkansas for Medical Sciences (UAMS) Institute for Digital Health & Innovation (IDHI) was one of 13 university teams selected as a North American regional finalist in the OpenCV Artificial Intelligence Competition.
Joseph Sanford, M.D.
There were nearly 1,400 entries worldwide in this competition, with 210 teams chosen as finalists across six geographic regions.
The IDHI team, named the Little Rock-ies, received the nod of approval to conduct non-invasive risk assessments of patient airways, with the use of specialized cameras that are equipped with spatial artificial intelligence, commonly referred to as spatial AI. As a finalist, the team won 10 cameras to complete the project.
“Respiratory illness has increased due to COVID-19, shining a light on the risks surrounding invasive airway procedures,” said institute director Joseph Sanford, M.D. “Using these spatial AI cameras, we are working on a solution that adapts existing assessments to better quantify difficult intubation risk, which can aid in selecting medications, equipment and techniques to safely secure a patient’s airway.”
Team members (from top left) Catherine Shoults, Pablo Trevino, Lori Wong, M.D., and Adria Abella Villafranca.
The spatial AI component of the camera provides human-level perception and real-time video to determine the distance and depth of the space inside of a patient’s airway as well as other features of the head and neck anatomy. Anesthesiologists currently use exam techniques that subjectively estimate these risks, meaning the assessment could vary from physician to physician.
In contrast, the team’s proposed method requires simply placing a camera in front of a patient’s open mouth, and the device will objectively calculate the measurements of the airway and provide an assessment score to determine the risks of invasive airway procedures, regardless of who the administering physician is.
Kevin Sexton, M.D.
As the team works to bring this new development to the forefront, one of the most significant factors of its success has been the contributions from UAMS students.
“Computer vision and artificial intelligence are new frontiers for health care, and it is thrilling to have the opportunity to develop this expertise,” said Catherine Shoults, a Ph.D. biomedical informatics student who also holds a master’s degree in public health. “As we continue to discuss these devices, we can continue to develop applications to advance health care.”
Alongside Sanford and Shoults, members of the team include Kevin Sexton, M.D., Lori Wong, M.D., Michael Cruz, Pablo Trevino and Adria Abella Villafranca.
Regional winners will be announced July 26.