Federated AI-Driven Digital Twins in the Healthcare Metaverse

Abstract

This chapter per the authors examines the integration of federated learning and digital twins to establish a resilient clinical intelligence ecosystem in the healthcare metaverse. While always-on sensing offers transformative potential, existing implementations often treat privacy, latency, and governance as secondary concerns. By synthesizing current literature, this chapter per the authors identifies a benchmarking gap where model metrics are prioritized over system-level evidence like communication overhead and energy costs. To address these deficiencies, authors propose a unified five-layer reference architecture treating trust and real-time synchronization as core requirements. This framework introduces the “protected twin” concept, utilizing differential privacy, secure aggregation, and audit trails to ensure distributed intelligence remains accountable and robust against poisoning. Ultimately, this chapter per the authors provides a strategic roadmap for moving toward deployable, clinically valid pipelines that balance innovation with patient safety.

Ähnliche Arbeiten