The infant is struggling to breathe, and you’re a medical student trying to save her life. The emergency room is a frenzy of beeping monitors, bustling nurses and a panicky mother. And nothing you’ve tried is working.
It feels real, but luckily it’s all virtual reality. ER doctors need to be prepared for anything, but they get little practice treating rare conditions they don’t face every day.
Startup AiSolve is taking the VR technology in gaming and using it to train doctors for rare conditions where lives are at risk.
“We want doctors to be better prepared,” said Shauna Heller, president for North America at U.K.-based AiSolve. She described her company’s GPU-powered solution during a talk Monday at the GPU Technology Conference, in San Jose.
Making it Real
Simulated training for doctors-to-be isn’t new. But students typically work with mannequins. Costs can run $430,000 a year for a single hospital to train surgeons on standard scenarios, according to AiSolve.
“It doesn’t feel real to med students or professionals, the costs are high and it’s inconvenient,” Heller said.
Most importantly, existing simulations are structured for one correct path and don’t allow students to make mistakes. “We offer a more natural way to learn, take chances and make mistakes in a safe environment,” she added.
VR Medical Simulation Pilot
AiSolve recently teamed with Children’s Hospital Los Angeles (CHLA) to build a pilot VR medical simulation for two rare but traumatic children’s conditions — seizures and anaphylactic shock, an extreme allergic reaction.
AiSolve used our GeForce GPUs to power its simulation and machine learning to incorporate expert knowledge in the system. CHLA doctors conceived the scenarios, which allow students to make decisions, conduct tests and evaluate their decisions based on responses by the virtual patient.
Wearing Oculus VR headsets, the students encounter a tense emergency room, with paramedics rattling off symptoms, nurses and technicians asking for a decision and a distraught mom praying for her baby’s survival. Facebook and Bioflight VR, a member of our Inception startup incubator, also contributed to the study.
AiSolve plans to incorporate deep learning in the next version of its product to enhance training and evaluation, Heller said.