*Presentation will be recorded.
Abstract:
This talk aims to show how Human-In-The-Loop (HITL) inspires, uplifts, and advances simulation methods in the context of clinical diagnosis and treatment. By selectively presenting three ongoing healthcare-simulation research led by Dr. Inki Kim, he intends to delineate the unique role of HITL simulation in conjunction with Machine Learning (ML), Human-Computer Interaction (HCI), and Model-Based Systems Engineering (MBSE). Prospectively, it is worthwhile to theorize the notion of simulation within those adjacent methodological domains (i.e., ML, HCI, MBSE), each one of which currently conceptualizes simulation on its own. Specifically, his first research project illustrates the novel use of mobile Mixed Reality (MR) technology for self-administered assessment of neurocognitive impairment, geared towards interoperable diagnosis and management of mild Traumatic Brain Injury (TBI). Here, HITL inspires the design of ‘adaptive’ agent in MR that deliberately interacts with the user, to disambiguate subtle signs/symptoms of neurocognitive impairment. For a work-in-progress, the cycles of action and perception of this agent within HITL will be guided via Reinforcement Learning. His second research illustrates viable ways that healthcare simulations expanded by Digital Twin (DT) support complicated decision-making. He will show a DT of Intensive Care Unit (ICU)–consisting of nurses, patients, electronic health records, and a tele-presence robot–all running on the simulated hospital facility at the Jump Simulation Center, Urbana. In MBSE, this DT allows for multi-model formalism as a groundwork to dictate the model update, as well as to build a co-simulation platform that leverages established simulation methods such as Discrete Event Simulation (DES) and multi-physics simulation. In HCI, the deployment of DT in eXtended Reality (XR) helps implement a new class of HITL that can potentially bridge people in simulation with those in the real-world. For his third project aimed at personalized migraine therapy, he will describe how HITL together with Generative Adversarial Network (GAN) model will be used to devise a non-invasive brain-stimulation treatment for chronic migraine patients. Ultimately, he envisions that healthcare simulations will expand to a wider array of healthcare works and patient-care contexts beyond traditional skill training, contributing to broader impacts on healthcare quality, cost, and access.
Bio:
Dr. Inki Kim is Assistant Director of Research at Health Care Engineering Systems Center (HCESC), a multidisciplinary research center at UIUC backed by $112.5 million in endowment support to foster collaboration between engineers and physicians. He is experienced with building various forms of interactive simulations including eXtended Reality (XR), brain-computer interaction, and haptic, and pursues through human-in-the-loop (HITL), the scientific discovery of human information-processing, particularly the cycle of perception and action. He has authored over 45 peer-reviewed scholarly articles and book sections in neuroergonomics, human-machine interaction, and modeling & simulation, and has been recognized several best papers by the Institute of Industrial and Systems Engineers (IISE), and the Human Factors and Ergonomics Society (HFES). Prior to joining UIUC in 2019, he was an Assistant Professor at University of Virginia, where he received the Research Innovation Award and the Graduate Medical Education Innovation Award. He received a PhD in Industrial Engineering from Pennsylvania State University in 2015. So far, he has led over fifteen research projects sponsored by the U.S. Air Force, Systems Engineering Research Center (SERC), Samsung, Hyundai Motors, LG, IBM Korea, and more recently, through the Jump ARCHES program. He is currently a Faculty fellow from the National Center for Supercomputing Applications (NCSA) at UIUC, and is an active member of the Society for Modeling and Simulation (SCS), the Institute of Electrical and Electronics Engineers (IEEE), and HFES.
