Privacy-preserving clinical data exchange using an AI-enabled zero-trust framework for Internet of Medical Things and cloud healthcare

The growing proliferation of the Internet of Things (IoMT) and cloud-based health care services has highlighted the demand for clinical data sharing in various heterogeneous platforms via secure and privacy-preserving communications. The paper presents an AID-based Zero-Trust solution combining behavioral anomaly detection, fuzzy trust evaluation, RL-based adaptive access control, federated learning, privacy-preserving computation and blockchain-based auditing. CNN–BiLSTM is used to detect behavioral threats, and fuzzy trust/policy optimization is applied to access decisions, with consideration for context. Collaborative analysis is achieved without directly sharing sensitive information by utilizing federated learning and secure computation, while blockchain offers tamper-resistant auditability. The proposed framework is evaluated by using data from intrusion detection systems of the health care sector to show its effectiveness for both classification and security purposes. The integrated architecture offers a holistic way of continuous verification, adaptive authorization, privacy protection, and accountable clinical data exchange in the distributed environment of IoMT and cloud-based healthcare.