DEMO
Anthony Nepomuceno, Doug Friedel, Ethan Man, Peter Grzywacz, Ariel Byun, Lydia Bramlet, Micah Bramlet, Victor Lu, Evan Cohen, Duo Wang, Avinash Gupta, Mae Vogel, Isaac Ryals, Shandra Jamison, Naveed Adoni, Caroline G.L. Cao, Sarita Adve, Cedric Dumas
High quality medical simulation training provides an important step between traditional didactic learning and the application of clinical skills at the patient bedside. Despite its importance, the number of training opportunities for medical students is often limited by staffing, budgetary, and time constraints. Extended Reality (XR) has the potential to provide additional training opportunities to medical students complementing the current medical simulation curriculum and the conventional use of high-fidelity mannikins, standardized patients, and replicated clinical environments. However, due to limitations of on-device computational power and network speeds, the quality of patient presence and dialogue can often leave much to be desired and, for certain simulation scenarios, be a barrier to facilitating learning outcomes. This work therefore aims to show methods to alleviate the computational and network bottleneck of XR devices to allow for higher quality patient animation and dialogue to support medical student learning outcomes for a focused patient history, physical examination, and communicating the diagnosis and care plan to the patient. This presentation will highlight the efforts to address these issues via offloaded rendering, procedural animations, and improvements to the conversational AI system. This is part of ongoing collaborative work with the NSF Agora Project leveraging cross-layer co-design to provide low-latency networking enabling remote medical collaboration and training.