Key-Embedded Privacy for Decentralized AI in Biomedical Omics

2026-03-30

AI Summary

This work addresses the challenge in biomedical omics research where stringent privacy regulations and data governance constraints hinder the sharing of raw data and the construction of representative cohorts. To overcome this, the authors propose INFL, a lightweight federated learning framework that innovatively embeds secret keys into an implicit neural representation architecture, yielding a coordinate-conditioned, plug-and-play module. This design enables seamless model aggregation across heterogeneous sites while preserving strong, controllable privacy guarantees without compromising model utility. INFL demonstrates superior performance across diverse tasks, including bulk proteomics classification, single-cell transcriptomic perturbation prediction, and spatial multi-omics clustering.

Abstract

The rapid adoption of data-driven methods in biomedicine has intensified concerns over privacy, governance, and regulation, limiting raw data sharing and hindering the assembly of representative cohorts for clinically relevant AI. This landscape necessitates practical, efficient privacy solutions, as cryptographic defenses often impose heavy overhead and differential privacy can degrade performance, leading to sub-optimal outcomes in real-world settings. Here, we present a lightweight federated learning method, INFL, based on Implicit Neural Representations that addresses these challenges. Our approach integrates plug-and-play, coordinate-conditioned modules into client models, embeds a secret key directly into the architecture, and supports seamless aggregation across heterogeneous sites. Across diverse biomedical omics tasks, including cohort-scale classification in bulk proteomics, regression for perturbation prediction in single-cell transcriptomics, and clustering in spatial transcriptomics and multi-omics with both public and private data, we demonstrate that INFL achieves strong, controllable privacy while maintaining utility, preserving the performance necessary for downstream scientific and clinical applications.

Problem

• privacy
• decentralized AI
• biomedical omics
• data sharing
• federated learning

Innovation

• federated learning
• implicit neural representations
• key-embedded privacy
• biomedical omics
• heterogeneous data aggregation