Chronic Obstructive Pulmonary Disease (COPD) is a leading cause of death in America. Visualizing COPD disease progression may help guide future care. We describe a method of generating a spiral atlas of temporal variation using a corpus of clinical notes to visualize COPD progressions. The atlas may help address a proxy problem – to predict the mortality of a given patient within a tolerance range, and may be useful in making recommendations for palliative care referral.

Learning Objective 1: Our objective is to assess the feasibility of automatically creating visualizations of COPD disease progression to death, then to identify and track patients with palliative care needs according to a patient’s probability of death.


Haohan Zhang (Presenter)
Fudan University

Chunlei Tang, Harvard Medical School
Joseph Plasek, University of Utah School of Medicine
Yun Xiong, Fudan University
Jing Ma, Harvard Medical School
Li Zhou, Harvard Medical School
David Bates, Harvard Medical School

Presentation Materials: