IEEE VIS 2024 Content: A Two-Phase Visualization System for Continuous Human-AI Collaboration in Sequelae Analysis and Modeling

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



Yang Ouyang - ShanghaiTech University, Shanghai, China. ShanghaiTech University, Shanghai, China

 Chenyang Zhang - University of Illinois at Urbana-Champaign, Champaign, United States. University of Illinois at Urbana-Champaign, Champaign, United States

 He Wang - ShanghaiTech University, Shanghai, China. ShanghaiTech University, Shanghai, China

 Tianle Ma - Zhongshan Hospital Fudan University, Shanghai, China. Zhongshan Hospital Fudan University, Shanghai, China

 Chang Jiang - Zhongshan Hospital Fudan University, Shanghai, China. Zhongshan Hospital Fudan University, Shanghai, China

 Yuheng Yan - Zhongshan Hospital Fudan University, Shanghai, China. Zhongshan Hospital Fudan University, Shanghai, China

 Zuoqin Yan - Zhongshan Hospital Fudan University, Shanghai, China. Zhongshan Hospital Fudan University, Shanghai, China

 Xiaojuan Ma - Hong Kong University of Science and Technology, Hong Kong, Hong Kong. Hong Kong University of Science and Technology, Hong Kong, Hong Kong

 Chuhan Shi - Southeast University, Nanjing, China. Southeast University, Nanjing, China

 Quan Li - ShanghaiTech University, Shanghai, China. ShanghaiTech University, Shanghai, China

 Room: Bayshore VI

2024-10-17T18:03:00ZGMT-0600Change your timezone on the schedule page
2024-10-17T18:03:00Z

Exemplar figure, described by caption below — System overview: Phase I includes (A) Cohort View for understanding drug event and disease progression relationships, (B) Patient Projection View to explore specific patient cohort characteristics, and (C) Medical Event View for detailed visualization of patient medical events. Phase II comprises (D) Modeling View for iterative AI model development and performance evaluation, and (E) Logs View for maintaining iteration records of models and associated data.

Fast forward

Keywords

Role Transfer, Hormone-related Medical Records, Visual Analytics, Machine Learning

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.