Toward a Human-Centered AI-assisted Colonoscopy System in Australia
Hsiang-Ting Chen, Yuan Zhang, Gustavo Carneiro, Rajvinder Singh
TL;DR
The paper addresses the gap between high-performing AI polyp detection and practical clinical adoption in Australia, using an field study of gastroenterologists to reveal the central role of user experience and workflow integration. It advocates a human-centered design approach, demonstrating that successful AI-assisted colonoscopy requires integrated UI/UX, robust quality assurance, and bias mitigation, not just advanced algorithms. Key contributions include insights into endoscopist needs, potential AI-supported QA and automation opportunities, and a call for co-design with industry and funding bodies. The work highlights the practical impact of integrating HCI into AI deployment to improve screening outcomes and equity in rural and urban settings alike.
Abstract
While AI-assisted colonoscopy promises improved colorectal cancer screening, its success relies on effective integration into clinical practice, not just algorithmic accuracy. This paper, based on an Australian field study (observations and gastroenterologist interviews), highlights a critical disconnect: current development prioritizes machine learning model performance, overlooking essential aspects of user interface design, workflow integration, and overall user experience. Industry interactions reveal a similar emphasis on data and algorithms. To realize AI's full potential, the HCI community must champion user-centered design, ensuring these systems are usable, support endoscopist expertise, and enhance patient outcomes.