Ozkar vision

The research section is a result of time spent on research and design of the project and is highly subject to change

0zkar will focus on addressing the needs of the ecosystem to help improve major bottlenecks in the healthcare industry: the increasing difficulty on collecting reliable, traceable data and capacity to collaborate on it.

The combination of zero knowledge proof and blockchain provides unique opportunities in building a more reliable infrastructure for the growing trend of decentralized clinical trials and decentralized studies. Increasing also the potential for valuable collection of Real World Data. New framework need also to emerge to better address the exponential growth of health digitalization ( genomics, medical devices, new sensors and IOT Equipment, image processing... ) and usage of AI in our healthcare systems.

The rise of personalized medicine is also increasing the pressure to better organized data collection and when needed avoid unecesary data transfer - which is a major promise of zero knowledge proof.

Ozkar development tools should allow to answer the needs of different stakeholder:

• patient: better access to innovation, transparency and trust about how their information is used. better guarantee from their contribution to research (ie: via escrow contracts, better data protection... )
• sponsors/private companies: decrease the compliance hassle by achieving the same results with less cybersecurity risks exposure. guarantee better data quality through zkp validation pipelines. Ensuring transparency towards authorities while avoiding IP exposure.
• researchers: better traceabilityto to their contribution for reseach and discovery. better indexing of metadata for their work to be identified and validated, facilitated collaboration with other research centers.
• healthcare professionals: transparency for their integration in protocols like decentralized clinical trials, studies or real world data collection (like quality of life) and better guarantee in the valuation of their work.

Ozkar 1.0 : web 2.5

The first months of research for the ozkar project have concluded to design a web2.5 approach which means that protocols will still need a trusted party but trust and transparency will still be highly improved by the usage of blockchain and zero knowledge proof. Due to the specificities of healthcare data, that still requires a lot of human expertise, verification, and processes that can't be covered by blockchain so far, we also believe it will still help building.

The architecture will rely on a ozkar oracle that needs to be deployed on a HIPAA compliant infrastructure( or equivalent). The ozkar-oracle is open source and can be deployed by any organization. Ozkar oracle/server will be responsible for several tasks:

• facilitating account creation and identity management until we are better equipped with decentralized identity.
• data validation alogrithms, including usage of zkml for data certification
• signing data for zkapp when and verifying signatures
• connecting to web2 data, signing and conversion to circuit compatible formats
• data storage (kept at minimum thanks to zkp)
• proof publishing

Data storage & encryption

Several storage solution will be provided in the long term. For the first phase , classic storage solution will be used with the ozkar server using a public key encryption in some cases which will allow to share data to a end user without allowing ozkar server to read but only to serve it to authorized consumers. In some other cases, ozkar server will need to run validation on the data which will require the server to access it, which is why oracle server needs to be running on HIPAA (or equivalent) compliant infrastructure. In a later stage, Threshold encryption with fhe / tfhe (from zama) will be introduced to enable collaborative studies between organization.

Overall design

The goal of ozkar is not to reinvent every aspect of data exchange. This is the reason why many struct will follow interfaces and work done by HL7 FHIR standard. The ozkar library will help manipulating this data according to this derivative standard and make it easy to use with ZKP and FHE.

Public components of the protocols, like data commitments will be published through the protokit framework hich allow to run its own app chain withouthaving to manage offchain storage for public data. Datasets will never be publicly available in the first phase but metadata can be published on ipfs to improve information retrieval.

Ozkar will also take a self service zkapp approach by allowing users to publish their own zkapp using a user friendly interface and escrow contracts. This has already been made possible by the MAC! project https://github.com/marekyggdrasil/mac

Ozkar server can also be asked to run specific validation algorithm against collected data. This where Ozkar will use ZKML techniques.

The goal is also to integrate as much as possible to existing tooling for data collection. A longer term goal is to test integration with Apple ResearchKit https://www.apple.com/lae/researchkit/ to provide great user experience.

#FHIR AND SMART ON FHIR

Ozkar aims to be compatible with the SMART on FHIR standard used by Apple. https://www.apple.com/healthcare/health-records/https://docs.smarthealthit.org/

You can use synthetic FHIR data generator for testing: https://github.com/synthetichealth/synthea

Healthcare data formats

In addition to FHIR data formats, we will investigate additional healthcare standards to be compatible with. One of them being OMOP, as this standard model is part a major european union initiative for healthcare data interoperability. https://www.ehden.eu/ OMOP data standard is also included in the plan from european commission for giving ownership of their health data to european citizen: https://health.ec.europa.eu/ehealth-digital-health-and-care/european-health-data-space_en . Allowing privacy friendly interoperability with blockchain and smart contracts using zk will unlock critical use cases.

Public databases

Ozkar could help develop and maintain highly valuable public database for healthcare by solving many issues within current systems: https://www.linkedin.com/feed/update/urn:li:activity:7152494078608183296/