This article was published on behalf of the author: Eugenio Gaeta, Platform manager of the Gatekeeper project.
The Gatekeeper platform is the heart of the Gatekeeper research and innovation project. In this article, we aim to deep dive into the platform to highlight the main features and describe its potential impact in helping to ensure healthier, more independent lives for the ageing population in Europe.
The Gatekeeper platform is designed to support both public and private applications within various geographical locations, that don’t need to be available locally most of the time, to facilitate data analysis and Artificial Intelligence applications for early detection and intervention in different healthcare contexts.
In the Gatekeeper architecture, pilot information systems are considered autonomous, collaborating application environments that share data and services on identical copies of the Gatekeeper platform. This kind of federated approach aims to avoid the unnecessary sharing of sensitive data while maintaining its interoperability at a technical, syntactic (meaning), semantic (grammar) and organisational level.
The main features of the Gatekeeper ecosystem include:
- High availability: by using new technologies for orchestration, replication of instances (e. g. OpenShift), fault tolerant systems and a dedicated infrastructure (Gatekeeper data centre), we will avoid loss of service by reducing or managing failures and minimising downtime.
- Scalable deployment: services and infrastructure of Gatekeeper platform instances are highly decoupled. Based on an «infrastructure as a code» approach, services will be able to run on any combination of infrastructure that meets minimum, high availability requirements.
- Interoperability of healthcare data and services: by using the HL7 FHIR standard, that defines how healthcare information can be exchanged between different computer systems regardless of how it is stored in those systems, the logical data model will be interoperable by default with all the healthcare centres of the project that adopt the FHIR standard. Gatekeeper has also made its FHIR Implementation Guide (FHIR-IG) public and shared it within the FHIR community.
- Interoperability with other domains: by using the Web of Things standard as a general service descriptor within Gatekeeper, it is also possible to link other domains, such as smart home or smart industry, to Gatekeeper services.
- Underpin data economy: the Gatekeeper Trust Authority, compliant with the International Data Space Association (IDSA) architecture, will contribute to the global data economy of the future European Health Data Space. With an initial IDSA compliance, Gatekeeper provides a link with Europe’s data strategy and the universal applicability of market-ready use cases.
- Moving processing closer to the data: since a Gatekeeper pilot platform instance contains its own online analytics data processing system, where possible, data analysis is first performed in the local environment and then aggregated in the federated platform. This approach will also ensure that sensitive data is not shared outside of the pilot environment.
- Optimised data analysis tools: since the data model is harmonised on the common FHIR data model described in the FHIR-IG, every platform instance within Gatekeeper provides standard pre-processing, aggregation, cleansing, and other data analytics services to accelerate the data science pipeline.
- Support federated learning approaches: the Gatekeeper platform makes it possible for AI algorithms to gain experience from a vast range of data located at different pilots. The federated learning approach enables several organisations to collaborate on the development of models, without needing to directly share sensitive clinical data with each other. In the same way, distributed map reduce paradigms will be enhanced to split the data mapping tasks into the specific pilot platform environment. Another important feature is the reproducibility of the analytics and AI pipelines to provide additional trust in and explanation of AI models.
With the Gatekeeper platform, we expect to generate an impact on the provision of AI services within public/private healthcare centres. The reusable approach and the validation studies of AI services developed within Gatekeeper, aim to improve the delivery of healthcare services for both patients and practitioners.