Developing an AI-driven multimodal approach to visualising resilient team performance: joint attentional engagement with gaze and speech in simulated emergency scenarios.

ABSTRACT:

Introduction

Healthcare team performance directly impacts the quality and safety of medical care. However, measuring the performance of teams is challenging and requires methodologies to investigate different contributing elements. This study proposes an AI(artificial intelligence)-driven multimodal approach to visualising gaze (ie, joint visual attention) and speech in medical team performance and examines how these might differ across medical expertise, using eye-trackers and our own automatic gaze annotation programme.

Method

Four simulation sessions, two in Japan and another two in the UK, were filmed with eye-trackers worn by a clinician and a nurse. In each site, one session was conducted with an experienced pair (UK_Ex and JP_Ex) and the other with a less experienced

SUBMITTER: Miyazaki A 

PROVIDER: S-EPMC12742140 | biostudies-literature | 2025 Dec

REPOSITORIES: biostudies-literature