A Three Pillar Data-Centric Framework for Enabling AI-Driven Healthcare Collaboration

Abstract

E-collaboration in healthcare has expanded through telehealth, remote monitoring, cloud EHRs, and connected diagnostics. However, despite unprecedented data availability, value extraction for collaborative workflows remains structurally limited. This chapter argues that data barriers—not modeling barriers—constitute the decisive bottleneck in AI-powered healthcare collaboration. While deep learning advances dominate discourse, algorithmic achievements depend on data that is ready, harmonized, semantically consistent, and identity-resolved, an assumption that is rarely met. This chapter presents a three-pillar framework: AI/Graph-Based Identity Resolution for unified patient representation across systems, AI/LLM-Driven Semantic Harmonization for consistent terminology, and Metadata Intelligence & AI-Native Data Fabric for machine-operable structure. Through analysis of data ecosystem components and real-world implementations including INPC, OHDSI, PCORnet, & eMERGE, this chapter demonstrates how addressing data infrastructure enables scalable AI-driven collaboration in healthcare.

Ähnliche Arbeiten