VPVitaliy Popov

Vitaliy Popov

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Vitaliy Popov has a total of 109 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    ReadGaze: Query the Gaze, Shape the Debrief

    by Haoting Gao, Kapotaksha Das, Liang Zhang, Jingying Wang, Mohamed Abouelenien, Michael A Cole, James M. Cooke, Chen Chen, Vitaliy Popov

  • Visual Attention is Not Memoryless: Higher-Order Sequence Modeling of TeamLead Gaze in VR Cardiac Arrest Simulation

    by Haoting Gao, Kapotaksha Das, Liang Zhang, Mohamed Abouelenien, Michael Cole, James Cooke, Vitaliy Popov

  • Meaningful Failure: Transforming Academic Structures and Incentives through Personalization in Engineering Education

    by Runu Das, Yeongji Lee, Tathyana Moratti, Spencer Sabatino, David Zabner, Mathew Baby, Katherine Becking, Herman Clements, Michael Brown, Allison Godwin, et al.

  • Open Access

    Towards Modeling Situational Awareness Through Visual Attention in Clinical Simulations

    by Haoting Gao, Kapotaksha Das, Mohamed Abouelenien, Michael Cole, James Cooke, Vitaliy Popov


  • Open Access

    Effects of reflection-through-action on collaborative clinical reasoning skills in an emergency dental simulation

    by Lauren Glowski, Nikolas Grotewold, Sarah Tomaka, Theodora E. Danciu, Vitaliy Popov

    Background: Simulation-based training is a central component of healthcare education. However, most current instructional designs emphasize post-simulation debriefing, while reflection during the simulation itself, termed reflection-through-action, remains understudied. Prior empirical evidence suggests that structured reflective pauses may enhance learners’ metacognitive regulation and diagnostic and intervention clinical reasoning, though whether these learning gains transfer to subsequent, novel clinical scenarios is unknown. This study examined whether reflection-through-action during an initial emergency simulation enhances the translation of technical skills and collaborative clinical reasoning practices in a subsequent emergency simulation among dental students. Methods: In this mixed-methods study, 110 dental students were randomized into teams and completed two sequential medical emergency simulations. Six experimental teams received structured reflection-through-action during the first simulation, consisting of a guided pause with written prompts and a structured PARSE (Pause–Analyze–Reflect–Strategize–Execute) debrief, while six control teams completed the simulation without reflective pauses. Outcomes accounted for prior knowledge and team familiarity and included expert-rated technical and non-technical performance, self-reported learning confidence, and behavioral process-level analyses using Epistemic Network Analysis (ENA) of team conversations during the second simulation. Results: Controlling for prior knowledge and team familiarity, no differences were seen in expert-rated technical or non-technical performance or in self-reported learning confidence between groups. However, ENA revealed important structural differences in the collaborative clinical reasoning displayed during the second simulation. Learners who engaged in reflective pauses demonstrated stronger connections between suggesting/requesting action, planning interventions, and intervening actions, indicating a greater connection between diagnostic reasoning and treatment management. In contrast, control teams showed weaker connections between planning and action, with conversations emphasizing procedural explanations, disconnected suggestions, and unfocused patient history questions. Conclusion: A structured reflective pause did not produce immediate differences on traditional performance metrics but revealed meaningful changes in collaborative clinical reasoning including how teams organized diagnostic, planning, and action during a subsequent emergency simulation. These findings suggest that reflective pauses reshape the structure of a team’s clinical reasoning and support near (contextually similar) and proximal (temporal) learning transfer. ENA offers a more sensitive and analytic approach than traditional performance-based outcomes for detecting differences in team-based collaborative clinical reasoning.

  • A Simulation-Based Root Cause Analysis of Pediatric Medication Dosing Errors in Emergency Medical Services

    by Bryan M. Harmer, John D. Hoyle, Lee Wells, Tycho Fredericks, Autumn Edwards, Adam Lecznar, Kathryn Christopher, Sue Dunwoody, Guan Hong, William Rantz, et al.

    Prehospital pediatric medication administration (PMA) carries a persistent dosing error rate of approximately 31%, despite efforts to mitigate these events. While the characteristics of these errors are well documented, underlying latent conditions in the emergency medical services (EMS) system contributing to these errors are not well understood. The purpose of this study is to identify the root causes within the EMS system that contribute to these errors. This study builds upon our previous mixed-methods simulation-based study that identified the frequency and characteristics of active errors associated with PMA. Eleven two-person EMS crews from two agencies in Michigan each completed three simulated pediatric emergencies requiring two doses of epinephrine, midazolam, or fentanyl. Simulations were recorded, and a physician and paramedic measured the accuracy of dosage administration. Post-simulation interviews were conducted to investigate actions associated with error and potential contributing factors. A multidisciplinary team analyzed equipment, protocols, interviews, observer notes, and participant comments through a modified Delphi method, identifying latent conditions using a Root Cause Analysis and Action framework and categorized them using a human factors framework. Latent conditions were identified and categorized after two rounds of the modified Delphi method. These conditions were found in all five tiers of the human factors framework. They included: 1) Individual Characteristics: limited clinician experience, knowledge, stress, and anxiety; 2) Nature of Work: high task complexity and limited teamwork; 3) Human-System Interface: cognitive aid and medical device usability issues; 4) Management: limited employee training and development; and 5) External Environment: protocol issues and a lack of assistive technology. These results highlight the critical systemic vulnerabilities underlying PMA dosing errors in EMS, supporting that errors result from systemic issues rather than solely from individual actions. By addressing these systemic weaknesses with comprehensive strategies, progress can be made in effectively resolving the ongoing issue of PMA dosing errors in EMS.


  • Assessing the Influence of a Statewide Dosing Reference Aid on Prehospital Pediatric Medication Dosing Errors: A Mixed-Methods Simulation-Based Investigation

    by Bryan M. Harmer, John D. Hoyle, Autumn Edwards, Tycho Fredericks, Lee Wells, Adam Lecznar, Kathryn Christopher, Sue Dunwoody, Guan Hong, William Rantz, et al.

    Errors occur in 31% of prehospital pediatric medication administrations (PMA) in Michigan despite the implementation of the MI-MEDIC pediatric dosing reference aid. Reference aids may not effectively target the causes of dosing errors and could introduce new types of error. Our study examines the frequency, magnitude, and immediate causes of PMA dosing errors in emergency medical services (EMS) after implementation of the MI-MEDIC dosing reference aid. We used a mixed method, mobile simulation-based approach to examine PMA dosing errors. Emergency medical services crews were recruited from a private EMS service and a fire-based EMS service in Michigan. Each crew completed three high-fidelity simulations (infant seizure, infant cardiac arrest and child burn) requiring two doses of midazolam, epinephrine, or fentanyl. Post-simulation interviews were conducted. Error rates and magnitudes were analyzed using descriptive statistics. A multidisciplinary team used a modified Delphi method, incorporating simulation observations, interview recordings, pictures of equipment that were used during the simulations, and protocols to categorize observed errors and reach consensus on their immediate causes. Eleven crews were recruited between both agencies, completing 33 simulations and 66 medication administrations, using the MI-MEDIC dosing reference aid. Dosing errors occurred in 19 of the 66 doses (28.8%), with 13 underdoses (68.4%) and six overdoses (31.6%). The median underdose was 0.50 (95% IQR: 0.20-0.56). The median overdose was 2.00 (95% IQR: 1.40-12.50). After two rounds, Delphi consensus categorized 13 slips (68.4%), six mistakes (31.6%), and no lapses. Immediate causes included crews obtaining the wrong weight (4/19, 21.0%), a large volume of air present in the syringe (6/19, 31.6%) and improper dilution of medications when following instructions on the MI-MEDIC reference aid (6/19, 31.6%). The remaining three involved epinephrine administration directly from the prefilled syringe system (2/19, 10.5%) and an intentional underdose (1/19, 5.3%). PMA dosing errors persist at a high rate with the use of the MI-MEDIC dosing reference aid. Immediate causes of error involve incorrect weight, administering air, and improper dilution, which tools like the MI-MEDIC dosing reference aid do not address. Further research is needed to develop comprehensive strategies addressing the active errors identified in our study.


  • Open Access

    AutoCLC: Towards Automated Assessment and Feedback on Closed-Loop Communication in Team-based Healthcare Simulation Training

    by Kaylee Yaxuan Li, Kunpeng Huang, Bryan Harmer, Casey Ducharme, Benjamin Heasman, Eleanor Falahee, James Cooke, Michael Cole, Alanson Sample, Vitaliy Popov

    This study introduces AutoCLC, an AI-powered system designed to assess and provide feedback on Closed-Loop Communication (CLC) in professional learning environments. CLC, where a sender’s Call-Out statement is acknowledged by the receiver’s Check-Back statement, is a critical safety protocol in high-reliability domains, including emergency medicine resuscitation teams. Existing methods for evaluating CLC lack quantifiable metrics and depend heavily on human observation. AutoCLC addresses these limitations by leveraging natural language processing and large language models to analyze audio recordings from Advanced Cardiovascular Life Support (ACLS) simulation training. The system identifies CLC instances, measures their frequency and rate per minute, and categorizes communications as effective, incomplete, or missed. Technical evaluations demonstrate that AutoCLC achieves 78.9% precision for identifying Call-Outs and 74.3% for Check-Backs, with a performance gap of only 5% compared to human annotations. A user study involving 11 cardiac arrest instructors across three training sites supported the need for automated CLC assessment. Instructors found AutoCLC reports valuable for quantifying CLC frequency and quality, as well as for providing actionable, example-based feedback. Participants rated AutoCLC highly, with a System Usability Scale score of 76.4%, reflecting above-average usability. This work represents a significant step toward developing scalable, data-driven feedback systems that enhance individual skills and team performance in high-reliability settings.


  • Development and Evaluation of a Multimodal Debriefing Dashboard for Virtual Reality Cardiac Arrest Simulation

    by Bryan Harmer, Jacob Erickson, Shane Ross, Dhruv Bhatnagar, Stephen Dowker, Hyunmin Park, Alanson P. Sample, Sarah Hartley, James Cooke, Michael Cole, et al.

    Introduction: Wearable sensing technologies incorporated into virtual reality (VR) headsets can provide a number of unique, previously hidden insights into individual and team learning and performance in complex environments. However, the most impactful and efficient mechanisms for conveying multimodal data (e.g., gaze, mental workload, etc) allowing for unique insights are not well understood. This study used a human-centered design process to develop and evaluate a multimodal debriefing dashboard. Methods: Twelve Advanced Cardiac Life Support (ACLS) Instructors completed VR cardiac arrest simulations and participated in interviews and design thinking workshops to identify dashboard features, leading to high-fidelity mock-ups and an interactive prototype. A second group of 6 ACLS instructors then evaluated the prototype by providing mock feedback on simulation performance with and without the dashboard in a pre-post format, followed by surveys assessing usability, acceptability, appropriateness, feasibility, trust, and perceived accuracy. Results: The final prototype dashboard features a video, a timeline of clinical actions, automatic error identification, mental workload, and visual attention data. It also generates closed-loop communication reports on a separate webpage. The evaluation study showed that participants provided performance analysis and feedback to students when using the dashboard, targeting specific (non)technical skills aligned with ACLS learning objectives. Participants rated the dashboard highly, with a System Usability Scale score of 88.9%, reflecting above-average usability. Conclusions: An interactive multimodal debriefing dashboard prototype was developed using a user-design framework with prototyping and iterative feedback. Use of the dashboard resulted in improved evaluation and debriefing practices targeting learning outcomes, with participants rating the dashboard favorably. This dashboard has the potential to enhance simulation-based learning by offering near real-time analytics that promote a deeper understanding of individual and team technical and nontechnical skill acquisition and mastery.