[Feature]: Homomorphic-Encryption

[JohnMoehrke]

Contact Details

No response

Is your feature request related to a problem? Please describe.

Should Privacy Enhancing Technology (PET) like Homomorphic-Encryption be discussed? Historically this kind of a solution was well beyond the scope of the IHE handbook, as there is no way to "profile" this in a standard way. Thus it would not have been useful for the intention of the handbook.

Describe the solution you'd like

No response

Describe alternatives you've considered

No response

Additional context

PET

Code of Conduct

• I agree to follow the IHE Code of Conduct

[JohnMoehrke]

Might possibly just introduce this, and express that it is a PET, but not something that we are Profiling.

[alan-zhang-22]

Gap: Homomorphic Encryption (HE) for Privacy-Preserving Collaborative Analysis. While multiple stakeholders have identified the need to support privacy-preserving collaborative analysis across clinical settings and health provider networks—scenarios where multiple organizations wish to perform joint computations (e.g., aggregate statistics, machine learning model training, cohort identification) without disclosing individual-level data—the current handbook does not provide guidance on homomorphic encryption techniques as a de-identification mechanism for this use case. Although HE offers a theoretically sound approach by enabling computation on encrypted data without decryption (thus eliminating the need to transmit or share plaintext records), integration of HE into the handbook has not been feasible due to two constraints: (1) External contributor participation: The technical expertise and real-world implementation experience necessary to properly specify HE techniques came primarily from practitioners outside the IHE membership structure, who lack formal standing in the workgroup's governance and decision-making processes. (2) Workgroup resource limitation: The de-identification workgroup lacks dedicated capacity to research, validate, and document HE techniques—including their cryptographic security assumptions, computational overhead, regulatory compliance implications, and practical deployment challenges—at the depth required for a normative contribution to the handbook. Impact of this gap: Organizations seeking to conduct collaborative secondary analysis across multiple health systems currently cannot reference this handbook for guidance on HE-based approaches, forcing them either to pursue ad-hoc implementations without established best practices, to rely on external cryptographic expertise that may not be coordinated with the broader de-identification governance framework, or to accept the alternative of centralizing data in a trusted secure enclave—with corresponding increase in data concentration risk. This limitation particularly affects research consortia, public health agencies conducting multi-site epidemiological studies, and healthcare networks pursuing federated learning initiatives, where neither traditional centralized de-identification nor pseudonymization alone fully satisfies both the privacy requirements and the analytical need for real-time computation on sensitive attributes across organizational boundaries.