When Lakeland Regional Health patients are admitted to the hospital, care teams must quickly assess their risk of falling and suffering a potentially serious injury. Traditionally, that process relies on clinical judgment and standard screening tools. Now, a group of Florida Polytechnic University capstone students is working to bring greater precision and data-driven insight to those critical decisions.
The seven-member interdisciplinary team of seniors is developing a tool that analyzes patient data to help assess fall risk. The tool will also recommend when one of the hospital’s limited number of high-tech monitoring devices would be most effective, helping care teams prioritize resources and improve patient safety.
“We want to use predictive modeling to find the probability of patients falling, given their data. From there, we can allocate the devices to those at highest risk of falling,” said Nolan Ross, an applied mathematics major. “Our computer science majors are designing an interactive UI (user interface) dashboard, and we’ll make additional improvements.”
At the center of the project is the AvaSure monitoring device for remote patient observation. It uses artificial-intelligence-powered computer vision and two-way audio and video to detect risky behaviors, such as someone getting out of bed. The remote observer can then speak directly to the patient or alert staff to provide immediate attention, allowing multiple patients to be monitored around the clock.
Factors that may increase a patient’s fall risk include medication side effects, cognitive impairment, muscle weakness and age.
“We are also creating an optimization model and analyzing benefits and costs to determine the financial value of how many AvaSure devices the hospital should have to minimize the fall risk,” said Hailey Bauer, a business analytics major with a concentration in logistics and supply chain management.
The team is composed of students majoring in applied mathematics, business analytics, data science and computer science.
“The jobs these essential workers are doing are very difficult and multifaceted, and we saw everything they have to manage all at once when we toured the hospital,” said Josephine Justice Johnson, a data science major. “If our project is able to alleviate just some of their daily stress, that’s extremely important.”
Team members said they were excited for the opportunity to apply their technical training to a project with real-world impact.
“We’re using real data. We have to clean the data, use the skills we were taught in class, and then train and test predictive models while we also look at how to present this to people who don’t necessarily know what we know,” Bauer said.
The team’s remaining members are data science major Melanie Najera, applied mathematics major Ryan Skornia, and computer science majors Danny Drysdale and Kelly Resetar.
Contact:
Lydia Guzmán
Director of Communications
863-874-8557
