- Peter Wonah Odey, Abdulazeez Ogunlade Babatunde and Myedan Nahum Eli
- DOI: 10.5281/zenodo.21967154
- SSR Journal of Artificial Intelligence (SSRJAI)
The digitization of Electronic Health
Records (EHRs) has significantly improved care coordination, but has
concurrently aggregated highly sensitive identity and clinical access data into
centralized repositories. These repositories serve as attractive targets for
unauthorized breach, insider misuse, and, with the rapid development of
cryptographically relevant quantum computers, future retrospective decryption
(commonly known as “harvest-now, decrypt-later” attacks). This paper
presents AQR-SSI, an Adaptive Quantum-Resilient Self-Sovereign Identity
framework designed to address these fundamental vulnerabilities. The
architecture seamlessly integrates: (i) a permissioned blockchain ledger for
decentralized identifiers (DIDs) and verifiable credentials (VCs); (ii)
zero-knowledge proofs (ZKPs) for unlinkable, selective disclosure of patient
attributes during Electronic Health Record (EHR) sharing; (iii) a hybrid
post-quantum cryptographic (PQC) suite combining NIST-standardized lattice- and
hash-based primitives (ML-DSA, Falcon, SPHINCS+, ML-KEM) with classical
algorithms to facilitate a secure migration period; and (iv) an explainable AI
(XAI) monitoring layer that continuously audits access requests and generates
human-readable explanations for anomaly flags via SHAP and LIME-based
attributions. We detailed the structural reference architecture. We also
described the actor and trust model, cryptographic protocol design,
smart-contract-based consent and revocation mechanisms, and conducted an
in-depth, threat-model-driven security analysis. Additionally, we provide an
illustrative, literature-grounded performance analysis estimating the
computational and storage overhead of the hybrid signature scheme relative to
classical elliptic-curve baselines, and discuss strict compliance alignment
under HIPAA and GDPR. The AQR-SSI framework is structurally adaptive: its
built-in cryptographic agility ensures algorithm suites can be easily swapped
as NIST standards evolve, while the XAI layer remains fully auditable as the
underlying anomaly detection models undergo periodic retraining. We conclude by
identifying key open problems to prioritize in empirical follow-up work,
specifically focusing on on-chain ZKP verification cost at scale, PQC signature
bloat, and the real-world usability of patient-held digital wallets.
Keywords:
Self-sovereign
identity, Blockchain, Zero-knowledge proofs, Explainable AI, Electronic health
records.
