Konnichiwa PsyErgo (CHI25, Yokohama)

Im Rahmen der Konferenz präsentierten Tobias Grundgeiger, Stephan Huber und Kilian Bahnsen in zwei Vorträgen und zwei Postern aktuelle Forschungsarbeiten der PsyErgo, von denen zwei als Honorable Mention ausgezeichnet wurden.

Abstract 1:

Reliable augmented reality (AR) cues can support the resumption of interrupted tasks. We investigated how sub-optimal AR cue reliability (100%, 86%, 64%, or no cue) affected users’ resumption performance and strategies. In a between-subjects experiment, 120 participants conducted a physical sorting task including interruptions, and we manipulated AR cue reliability (i.e., the AR cue was present or absent at the end of interruptions). In trials with AR cue, performance with 86% and 64% reliable AR cues was as well as with 100% reliable cues. In trials without AR cue, performance with sub-optimal AR cue reliability declined but was still better than with no cue. Cue reliability affected task resumption strategies of the 86% (slow but no increase in errors) and the 64% (fast but increase in errors) reliability groups differently. Our results extend reliability research to interruptions and the observed efficiency-thoroughness trade-offs in resumption strategies provide insight for design.

Abstract 2:

In addition to ensuring patient safety during anesthetic inductions, anesthesiologists must document clinical interventions and administer drugs. This is a time-consuming and low priority task, which harms the documentation quality of anesthetic protocols. In this case study, we demonstrate how speech-based artificial intelligence (AI) assistants that leverage closed-loop communication can increase documentation quality. An evaluation in 40 scenarios in a medical high-fidelity simulator indicated that the AI documentation assistant facilitated earlier data entry and increased documentation precision. However, despite the objective advantages for data quality and patient safety, anesthesiologists experienced a higher temporal demand with the system. With this study, we contribute qualitative insights of how the AI documentation assistant benefited anesthesiologists’ work style and affected their interactions within the team. Future research should aim to design AI assistants that enforce communication clarity while considering their impact on team dynamics.

Abstract 3:

When referring to the role of newly proposed clinical applications of artificial intelligence (AI), recent work inflationary uses the term Human-AI Team. However, the roles foreseen for AI systems within teams remain unclear. We systematically reviewed the literature on AI deployment in anesthesiology and found that most AI system papers only describe algorithms. We identified 57 interactive systems and assigned six roles based on described behavior, tasks, and interactions. While the most prevalent role was task completer, some AI systems also served their team as problem solvers, evaluators, task motivators, or even teamwork support and team leaders. We contribute (1) a classification system for team roles, behaviors, tasks, and interactions of AI team members and (2) an overview of AI systems’ team roles in anesthesiology. We conclude that (3) AI systems’ intended social roles within teams need to be more consciously reflected, shaped and clearly communicated to meet healthcare standards.

Abstract 4:

Recent advancements in artificial intelligence have sparked discussions on how clinical decision-making can be supported. New clinical decision support systems (CDSSs) have been developed and evaluated through workshops and interviews. However, limited research exists on how CDSSs affect decision-making as it unfolds, particularly in settings such as acute care, where decisions are made collaboratively under time pressure and uncertainty. Using a mixed-method study, we explored the impact of a CDSS on decision-making in anesthetic teams during simulated operating room crises. Fourteen anesthetic teams participated in high-fidelity simulations, half using a CDSS prototype for comparative analysis. Qualitative findings from conversation analysis and quantitative results on decision-making efficiency and workload revealed that the CDSS changed team structure, communication, and diagnostic processes. It homogenized decision-making, empowered nursing staff, and introduced friction between analytical and intuitive thinking. We discuss whether these changes are beneficial or detrimental and offer insights to guide future CDSS design.

Referenz:

• Kilian L Bahnsen, Emma Dischinger, and Tobias Grundgeiger. 2025. AR Cue Reliability for Interrupted Task Resumption Affects Users' Resumption Strategies and Performance. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI '25). Association for Computing Machinery, New York, NY, USA, Article 1224, 1–12. https://doi.org/10.1145/3706598.3713685
• Stephan Huber, Ronja Fricke, Caroline Pütz, Lennart Baumeister, Christina Dilling, Oliver Happel, Simon Ottenhaus, Anja Nagel, Matthias Dunkelberg, and Tobias Grundgeiger. 2025. Evaluating an AI Documentation Assistant for Anesthesiology Teams. In Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA '25). Association for Computing Machinery, New York, NY, USA, Article 672, 1–8. https://doi.org/10.1145/3706599.3706658
• Stephan Huber, Lea Weppert, Lennart Baumeister, Oliver Happel, and Tobias Grundgeiger. 2025. Team Roles of Artificial Intelligence in Anesthesiology – A Scoping Review. In Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA '25). Association for Computing Machinery, New York, NY, USA, Article 508, 1–13. https://doi.org/10.1145/3706599.3720186
• Sara Wolf, Tobias Grundgeiger, Raphael Zähringer, Lora Shishkova, Franzisca Maas, Christina Dilling, and Oliver Happel. 2025. How a Clinical Decision Support System Changed the Diagnosis Process: Insights from an Experimental Mixed-Method Study in a Full-Scale Anesthesiology Simulation. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI '25). Association for Computing Machinery, New York, NY, USA, Article 257, 1–23. https://doi.org/10.1145/3706598.3713372