TAPESTRY: Temporal Analytics Platform for Exploratory Scientific and Translational Research inquirY

Poster Number: P43
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Informatics Implementation, Information Visualization, User-centered Design Methods
Primary Track: Applications
Programmatic Theme: Clinical Informatics

TAPESTRY is a modular, generalizable toolkit designed for clinical scientists and informatics engineers to rapidly develop temporally oriented patient datasets. This allows them to study disease progression, treatment efficacy, and longitudinal patient care. The platform achieves temporal orientation by adopting a universal patient:time key, allowing the inclusion of patient data from any system. Each TAPESTRY instance is envisioned as a component in a larger "data fabric," offering a scalable ability to combine and view longitudinal clinical records, each specialized to a specific domain. The TAPESTRY interface harmonizes clinical experts and developers, providing a continuous feedback loop through embedded annotation workflows. This toolbox-driven approach can provide a new, technically supported platform to enhance collaboration and expertise sharing throughout an institution. TAPESTRY has been implemented in 3 clinical domains: ICU, Cardiology, and Cancer, providing a richer ability to stratify patient risk by shifting focus to time at risk. The platform is an effective tool for fine-grained cohort identification and selection, disease state change tracking, and rich comparative analyses. Plans are underway to extend the platform with greater group identification and group comparative analytics capabilities via the implementation of advanced techniques for event detection, forecasting, clustering, and phenotyping of time series profiles.

Speaker(s):
Albert Riedl, M.S.
UC Davis Health System

Author(s):
Nicholas Anderson, PhD - University of California, Davis;