International Digital Health & AI Research Collaborative

PathFinders provide the framework for I-DAIR’s projects portfolio and an end-to-end approach to the global research ecosystem for digital health, from infrastructure and tools to benchmarking and governance. Current projects include the following:

• Global Research Map of Digital Health and AI (GRM)
• Real-Time Epidemiology and Dashboard (RTED)
• Digital Innovations in Health Systems (DIHS)
• Trusted Research Infrastructure (TRI)
• Capacity Development Network (CDN)

–    Governance for AI and Data for Health

The Global Research Map of Digital Health and AI

I-DAIR’s Global Research Map of Digital Health and AI (GRM) is an interactive tool designed to provide a national, regional, and global outlook on digital health and AI research and innovation. The map is based on a new methodology, which combines a technological approach based on AI (natural language processing) to comb through open-source data from 2011 to 2020, along with qualitative surveys across nine regions of the globe, to generate digital health and AI R&D terms and identify trends and trendsetters.

It includes a powerful visualization tool on the patent and development landscape, research publications, and R&D trends in digital health from different angles, including thematic, geographical, and chronological. The level of maturity between different national strategies can also be explored within the initial set of 23 countries. Governments can benchmark themselves with their peers using a set of neutral criteria. They can analyze current strengths while prioritizing investments in future capabilities and then monitor the impact of their strategies on research and innovation annually. To access the maps and the GRM resources, visit https://grm.i-dair.org.

Real Time Epidemiology and Dashboard

The Real Time Epidemiology & Dashboard (RTED) PathFinder facilitates digitally enabled pandemic surveillance and response. Under this PathFinder, I-DAIR is elaborating with its partners a global, science-based, data-driven, neutral, and trusted collective capacity, which leverages citizen science to improve the quality of local and national responses throughout the continuum of pandemic phases.

The proposed global end-to-end pandemic preparedness and response scheme will adopt a modular architecture that will allow responses to be ramped up or down as the pandemic progresses. Countries can adopt what is feasible according to their urgency and grow the scheme as their capabilities increase, and more resources become available.

Under such a scheme, a myriad of innovative data sources and digital technologies can come together locally, regionally, and globally within a neutral digital distributed infrastructure to provide a robust and trusted evidence base to drive responsive public health decision-making.

Digital Innovations in Health Systems

Within the DIHS PathFinder, the aim is to develop digital or digitally enabled innovations that help strengthen health systems and improve access to essential health services as part of the push towards the goal of universal health coverage by 2030. I-DAIR is currently exploring digital innovations in three areas: primary health care and preventive health through the Open Health (OH) scheme; Maternal and Child Health (MCH), with the MOTHER project; and Antimicrobial Resistance (AMR), with AI-enabled antibiotic prescribing decision support systems.

OH is a project in partnership with the governments of Punjab and Nagaland in India. It aims to go beyond the current telemedicine paradigm towards a more open and community-centered healthcare approach, where a distributed digital infrastructure, combined with locally available clinical and medical knowledge, provides a continuum of care from promotion, prevention, and early diagnosis to treatment.

Trusted Research Infrastructure

In essence, the TRI is a distributed hardware infrastructure for the (con)federated use of data and AI that can scale independently of the current infrastructure areas dominated by the tech giants by focusing on small states and low and middle-income countries (LMICs) as well as innovators and researchers (rather than businesses and individual customers).

The TRI will support the unique characteristics and needs of health data and clinical trials data that are not well-supported today by the major cloud hosting companies. Simultaneously, it will put theories to the test by creating an environment for hands-on practice as well as prototyping and demo opportunities. Moreover, the TRI will provide reliable capacity from off-the-shelf to entirely do-it-yourself AI and machine learning tools with full data sovereignty, thereby speeding up the development of AI capacity within I-DAIR’s partner institutions.

Governance and AI

With a recently obtained grant from the Canadian International Development Research Centre (IDRC), I-DAIR will collaborate through its research hubs and partners in LMICs to develop a better understanding of how to implement responsible, gender-responsive AI across a wide variety of contexts in two specific domains: (1) epidemic/pandemic preparedness and response; and (2) sexual, reproductive, and maternal health.

The project will integrate principles into practice by co-creating and validating with IDRC and its research hubs and partners a governance mechanism, including processes and tools to apply existing international principles, mechanisms, and norms in LMIC contexts. It will include stakeholder engagement mechanisms and support the exchange of governance insights and assets.

The governance mechanism will be made available to the global community of AI researchers and implementers. Moreover, by increasing the number, diversity, and contextual specificity of trusted governance tools, international regulatory discussions will be more nuanced and enriched by perspectives from LMICs.