A Two-Phase Visualization System for Continuous Human-AI Collaboration in Sequelae Analysis and Modeling

TL;DR

This paper addresses sustaining effective human-AI collaboration in complex healthcare data by introducing a two-phase visualization framework for sequelae analysis. Phase I enables Human-Led, AI-Assisted Retrospective Analysis with Cohort, Projection, and Event views to surface risk factors, while Phase II supports AI-Mediated, Human-Reviewed Iterative Modeling using backend experiments (RandomForest + SHAP) and a Modeling/Logs interface for iterative refinement. Expert feedback indicates improved analysis efficiency and clearer human-AI collaboration, though concerns about data quality and AI overreliance remain. The framework is extensible to other domains and emphasizes interaction-driven co-design and interpretability to guide practical healthcare analytics.

Abstract

In healthcare, AI techniques are widely used for tasks like risk assessment and anomaly detection. Despite AI's potential as a valuable assistant, its role in complex medical data analysis often oversimplifies human-AI collaboration dynamics. To address this, we collaborated with a local hospital, engaging six physicians and one data scientist in a formative study. From this collaboration, we propose a framework integrating two-phase interactive visualization systems: one for Human-Led, AI-Assisted Retrospective Analysis and another for AI-Mediated, Human-Reviewed Iterative Modeling. This framework aims to enhance understanding and discussion around effective human-AI collaboration in healthcare.

Figures (1)

• Figure 1: The two-phase framework overview.