DGUARD

DGUARD offers a comprehensive open-source framework designed to facilitate secure data sharing while prioritizing user privacy and security. Our framework incorporates essential features such as data segmentation, provenance tracking, and robust auditability. Its modular design allows for easy integration via APIs or SDKs, and includes the following key components:

Consent Management System: Utilizing self-sovereign digital identity authentication methods, this system ensures transparent and user-centric control over data sharing permissions.
Privacy-Preserving Authentication Toolkit: Leveraging zero-knowledge proofs (ZKP), this toolkit enables data anonymization while maintaining segmentation capabilities, safeguarding user privacy.
Proxy Re-encryption Scheme: Employing advanced encryption techniques, this component guarantees end-to-end data privacy irrespective of the data controller or data transit path.
Blockchain Notarized Audit Trail: Utilizing blockchain technology, this feature provides a tamper-proof audit trail for data transactions, ensuring traceability, non-repudiation, and accountability at every step.

By integrating these components, our framework offers a robust solution for organizations seeking to share data securely while upholding the highest standards of privacy and accountability. The project will be piloted in a real environment to securely share dermatology data for federated learning for early melanoma detection in controlled set of 30 clinics, 100 patients and 20 private practitioners.

Motivation for the project: dGUARD framework of tools aims to revolutionize how data-sharing procedures are conducted, with a primary focus on enhancing consents management, ensuring data security, privacy, and anonymity and providing robust data-sharing processes with verifiable traceability to guarantee accountability.
Generic use case description: The use case we will be integrated with the Top Doctors Platform to be piloted in a real scenario to securely share dermatology data for federated learning for the early detection of melanoma. We will employ Federated Learning and public blockchain technology to securely federate images, analytics, and diagnostic information, including BRCA2+ mutation indicators. Data will be securely stored in the interoperable health repository. Only data meeting specific ethical and medical criteria will be selected, employing ZK Auth and self-sovereign identity for verification.
Essential functionalities: The objective of this project is to build, deploy, test, and provide an end-to-end framework for addresses the challenges of ensuring privacy, accuracy, consent management and traceability of data and solves them by proposing solutions on 1) Lack of control over access to personal information. 2) Lack of anonymity and privacy. 3) Lack of data traceability and accountability. 4) Lack of data security, privacy and integrity in data storage and shared data process. dGUARD offers an innovative solution to address the challenges mentioned above. We are committed to delivering a reusable solution that harnesses the power of self-sovereign digital identity (SSI), Zero-Knowledge Proofs (ZKP) for authentications, proxy re-encryption, and blockchain notarization. The system will capacitate any data sharing platform to with: (F1) E-consent management: Simple interface, granular controls, easy opt-in/out. Provide user-friendly & robust electronic consent acceptance that enhance the whole process privacy management security & usability. (F2) Privacy-first data-sharing: Anonymization, minimal data, default privacy, user authentication and secure transmission. Guarantee the anonymity, data minimization and privacy of transmitted data by default. Remain in control of data sharing. (F3) Verifiable traceability: Complete verifiable logs & audits, for robust accountability. Facilitate a transparent and verifiable record of all data exchange processes executed thus enabling traceability over data handling
How these functionalities can be integrated within the software ecosystem: The privacy-by-design components have been created to seamlessly integrate with various systems through Open APIs and Integration Toolkits built on SDKs. Optionally you can utilize these components to deploy user-friendly front-end interfaces that facilitate easy access and interaction for users.
Gap being addressed: The addressed gap concerns (1) Lack of control post-sharing and challenges with e-consent ; (2) Insufficient privacy measures with the controller as policy executor and authenticator. (3) Lack of accountability, traceability, and transparency in data sharing processes.
Expected benefits achieved with the novel technology building blocks: We expect to develop a fully functional open-source framework that enables the management of users consents and privacy-preserving data sharing (anonymity and privacy) among parties whilst maintaining visibility and verifiability of all stakeholders' interactions.
Potential demonstration scenario: The envisioned demonstration scenario entails a controlled real-world setting where dermatology data is securely shared for federated learning aimed at early melanoma detection. This scenario involves 30 clinics, 100 patients, and 20 private practitioners.