Blockchain Governed Federated Analytics for Auditable Medical Data Analysis

Abstract

Medical data sharing across institutions is frequently constrained by legal, regulatory, and organizational barriers, making centralized aggregation impractical in many healthcare settings. This paper presents a proof-of-concept platform for governed, collaborative medical analysis based on federated analytics and blockchain-backed audit receipts. The platform combines Hospital Agents, Dataset Descriptors, consent policies, access requests, federated analytical jobs, and an off-chain computation and on-chain recording model. Raw data remain within the participating institution, while the central orchestrator coordinates approved analytical tasks and records selected governance receipts and consent states on-chain. The prototype is evaluated through measurements of governance overhead and comparison with IBM SPSS for descriptive statistics and Pearson correlation. The results indicate that the proposed workflow can reproduce selected exploratory statistical analyses without centralizing the underlying dataset, while providing tamper evident receipts for key governance actions. The current implementation remains a controlled proof-of-concept and does not address adversarial robustness, verification of off-chain computations, or deployment in production hospital infrastructures.