临床环境中符合HIPAA的AI部署模式：隐私保护技术与治理控制
HIPAA-Compliant AI Deployment Patterns in Clinical Settings: Privacy-Preserving Techniques and Governance Controls
浏览论文内容
中文总结 AI 辅助
本文提出五种可复用的临床AI部署模式，联合应对HIPAA隐私与安全规则，并提供威胁模型与模式选择决策程序，将监管义务转化为可验证的架构控制。
中文摘要 AI 辅助
人工智能（AI）正从研究原型走向临床工作流程，然而，处理受保护健康信息（PHI）的AI系统的部署仍受到美国《健康保险携带与责任法案》（HIPAA）以及缺乏满足该法案的共享工程指南的制约。现有工作往往将隐私保护机器学习与监管治理视为分离的问题，使从业者缺乏从架构决策到合规义务的具体映射。本文贡献了一个包含五种可复用临床AI部署模式的目录，这些模式共同应对HIPAA隐私与安全规则：去标识化分析区、跨实体联邦学习、可信执行环境内的机密推理、最小化PHI的检索增强临床助手，以及用于开发的合成数据沙箱。每种模式均根据上下文、主导力量、结构和残余风险进行规范，并映射到HIPAA安全规则的管理、物理和技术保障措施，以及NIST AI风险管理框架和ISO/IEC 42001。我们进一步提出了临床AI的威胁模型和基于数据敏感性、信任边界和延迟要求选择模式的决策程序。目标是给安全工程师、合规负责人和临床信息学专家提供一种共同词汇，将抽象的监管职责转化为可验证的架构控制。
英文摘要
Artificial intelligence (AI) is moving from research prototypes into clinical workflows, yet the deployment of AI systems that process protected health information (PHI) remains constrained by the U.S. Health Insurance Portability and Accountability Act (HIPAA) and by the absence of shared engineering guidance for satisfying it. Existing work tends to treat privacy-preserving machine learning and regulatory governance as separate concerns, leaving practitioners without a concrete mapping from architectural decisions to compliance obligations. This paper contributes a catalog of five reusable deployment patterns for clinical AI that jointly address the HIPAA Privacy and Security Rules: a de-identified analytics zone, cross-entity federated learning, confidential inference within trusted execution environments, a PHI-minimizing retrieval-augmented clinical assistant, and a synthetic data sandbox for development. Each pattern is specified in terms of context, governing forces, structure, and residual risk, and is mapped to the administrative, physical, and technical safeguards of the HIPAA Security Rule as well as to the NIST AI Risk Management Framework and ISO/IEC~42001. We further present a threat model for clinical AI and a decision procedure for selecting patterns based on data sensitivity, trust boundaries, and latency requirements. The goal is to give security engineers, compliance owners, and clinical informaticists a common vocabulary that turns abstract regulatory duties into verifiable architectural controls.
