Spotlight Projects

Image
XR/AR/VR Training Solutions
Designed in collaboration with medical experts, our immersive system offers a highly realistic training environment for students learning life-saving emergency procedures. Virtual patients respond dynamically to spoken commands and questions, allowing trainees to diagnose conditions and perform procedures in real time, all governed by a sophisticated physiology engine that simulates injuries as they evolve. In addition, we developed a complementary instructor back-end that enables real-time adjustment of injury severity. We combined this with a user-friendly graphical interface that displays critical information at a glance—empowering educators to tailor training scenarios on the fly. 
Applications
TRIAGE
This study investigated the impact of cognitive load on medics operating in high-stress field conditions. Using the HoloLens 2, virtual wounded soldiers are placed into real-world environments, allowing medics to practice diagnosing and treating injuries while facing an array of carefully designed stress-inducing distractors. These distractions, such as thick smoke, the cries of injured people, and blaring building alarms, replicate the chaos of battlefield scenarios. The study aims to better understand how cognitive load affects performance by simulating these high-pressure environments, ultimately improving training methodologies and enhancing the readiness of medical personnel in critical situations. 
Prolonged Field Care

Single and multi-patient trauma scenarios, covering point of injury treatment through extended care. It features an after-action review and scoring system based on the TCCC MARCH protocol. This system provides feedback on what the user did correctly and areas for improvement, along with optional suggestions for enhancing performance.

It includes a built-in scenario editor to create your own custom scenarios and is available on multiple platforms, including tablets, PC, and Mac. 

PrESH
VR Project with UNC to develop high-quality, interactive, and realistic digital therapeutic simulations designed for teenage patients to practice cognitive behavioral therapy skills via natural language processing (NLP) and facilitate skill learning in a clinical setting to improve school reintegration experience.  
Abdomen Trainer
VR anatomy trainer demonstration. Highly detailed digital training manikin with deformable organs based upon their physical properties, and the labeling of organs. 
Image
Image
Image

Medical Training Solutions

Instructors can create scenarios with up to three patients with varying degrees of injury severity. They can also schedule additional complications to trigger after a specified time. Further injuries can be configured based on student actions, external factors, or direct instructor input, offering a customizable training experience that can be predetermined or dynamic.  

Powered by an open-source physiology engine, the system delivers real-time, medically accurate vital signs and corresponding audio-visual cues to simulate realistic patient responses. Its NLP-enabled dialogue feature allows patients to interact with students by responding based on injury severity, treatments applied, or repeated questions, enhancing the realism of each scenario. 

The system also includes a comprehensive after-action review that provides detailed feedback on student performance using the TCCC standards, outlining correct actions and mistakes, and a percentage score to measure overall proficiency. 

Image
Image
Image

AI Soft Skills Training

We developed a highly customizable training platform that allows users to create tailored scenarios based on specific learning objectives. After selecting from multiple locations and virtual characters, instructors can define unique goals and add supporting information like resumes or medical records to add realistic details to their scenarios. To get started faster, they can even duplicate and modify existing scenarios to fit different needs. The platform supports conversation trees for more interactive learning experiences, giving users a wide range of options for developing realistic, engaging simulations. 

Our solution features realistic virtual characters with phoneme-based lip synchronization, natural eye movement, and eye contact, enhancing immersion. Thanks to our advanced NLP and semantic AI mapping, users can interact directly with the characters with their natural voice all within a 3D environment from their browser.  

The system also includes a comprehensive after-action review, which provides performance scores, feedback on what was done correctly or incorrectly, and even the ability to replay scenarios for in-depth analysis and improvement. 

Image
Image
Image
Image
Image
Image