Enhancing Accuracy, Time Spent, and Ubiquity in Critical Healthcare Delineation via Cross-Device Contouring

Abstract

Improving accuracy, time spent, and ubiquity of delineation has been a long-standing design aim, yet many HCI works have overlooked high-stakes and complex healthcare annotation. We explore contouring, a critical workflow aimed at identifying and segmenting tumors, usually performed on immobile desktop computers in clinics, in which limited support for mobile access leads to prolonged and subpar treatment planning. Following interviews and think-aloud studies (N=10 physicians), we report key contouring behaviors, and later design a novel cross-device prototype that enables contouring on everyday touch devices. We compared contouring via desktop and touch in a lab study (N=8 residents) and found that mobile phones not only yielded similar accuracy, but also took significantly less time. Our results point to three broad design guidelines for cross-device solutions deployed within standalone healthcare workflows, and highlight how incorporating different device and input modalities can improve treatment delivery in distributed healthcare environments.

Matin Yarmand

Ph.D. Candidate

Chen Chen

Ph.D. Candidate

Nadir Weibel

Professor of Computer Science and Engineering