IDAIR

Our PathFinders

Our PathFinders provide a space for innovation to tackle global health issues

 

The PathFinders will help build the foundation for I-DAIR to become a neutral and trusted platform for global collaboration on digital health and Artificial Intelligence for solving health problems.

An I-DAIR research project could explore and contribute to several PathFinders at once. Our partners in these projects benefit from the scientific learnings and technologies developed through the PathFinders, which create a digital ecosystem for developing and sharing global public goods.

Through expert consultations, I-DAIR has identified ten PathFinders. For the incubation phase (running until 2022), we have chosen to focus on six PathFinders out of the original ten (highlighted):

Global Research Map

What global health problems does the GRM engage with?

To date, no comprehensive global research map or overview of research activities on Artificial Intelligence and Digital Health is available.

Our Global Research Map is essential to provide situational awareness to regions, countries and multilateral organisations for developing their own strategies and digital health activities.

There are several ways in which our map can enable solutions for global health problems:

  • It provides a floor for discussions in global health policy bodies such as the World Health Assembly;
  • Partners can use it together with the report delivered by the joint Lancet/Financial Times Commission to identify policy and governance gaps for the achievement of universal health care and SDG3;
  • It enables I-DAIR and donor agencies to identify gaps and potential areas of extension and interconnection.

What approaches will this PathFinder take to tackle these problems?

I-DAIR will undertake a taxonomy development and mapping exercise, which includes the following:

  • An inclusive process of consultations which gathers information on emerging R&D efforts in diverse geographies of innovations;
  • An AI-based stakeholder mapping tool used in parallel with the consultations to tease out the latest research and technology development trends;
  • A visualization tool for presenting the GRM.

Real Time Epidemiology & Dashboards- RTED

What global health problems does the RTED PF engage with?

This PathFinder explores collaboration challenges and facilitators for the digitally-enabled prediction of the outbreak, spread and management of epidemics.

The goal is to combine data with data analysis and machine learning tools to create a data-ecosystem for real-time epidemiology and the prediction of pandemics. The real time epidemiology & dashboards PathFinder also explores digital tools for public health responses and decision making dashboards.

What approaches will this PathFinder take to tackle these problems?

During the COVID-19 pandemic, I-DAIR together with its partners (government, academia and private companies), have started exploring the public health response to COVID-19 from a citizen-centric and science-based perspective. The focus has been on developing modeling diversity, on integrating molecular science data into mobility and other types of epi-data, and on bringing narratives or stories as proxy-variables in place of missing numerical data. Our goal is to build on these pilots to develop more advanced AI models and digital tools to inform and power policy- and decision-makers in real-time on epidemic spread and pathogen evolution.

Data Architectures & Data Interoperability – DADI

What global health problems will the DADI PF engage with?

Fragmented datasets, zero-sum views of data use, lack of access to public or private datasets and short stubs of digital diagnostic data are inhibiting large scale deployment of data and AI for health. The digital interoperability problem sits atop analog interoperability problems coming from impervious structures serving narrow mandates, specific diseases and organ centric health silos.

To develop the next generation of digital health solutions we require:

  • A degree of interoperability within and across data architectures;
  • Clear ways to make use of specific health data sets in a globally distributed manner;
  • Research on governance and interoperability issues as the quality and autonomy of digital diagnostic flows is stepped up.

What approaches will this PathFinder take to tackle these problems?

We believe that there is no silver bullet for interoperability and analog interoperability is as important as digital interoperability. I-DAIR plans to survey various levels of interoperability, test interoperability solutions in diverse geographies with diagnostic flows from concrete use cases and develop 3-4 model approaches to interoperable digital infrastructures for health.

These model pathways will be developed in I-DAIR hubs based on the highest regulatory and data protection standards and full respect for patient privacy and health worker agency. Private sector involvement particularly at the front end of diagnostic data flows will be prioritised. This will help develop shared use of health data and algorithms for innovative digital health solutions.

The models act as a stimulus to building the health data infrastructure of the future and identify the right incentives for private-public data integration and innovative business models. They also allow citizens to participate in addressing questions around data ownership and privacy.

Benchmarking

What global health problems will the Benchmarking PF engage with?

How do we know it works? The challenge we face when responsibly scaling digital health innovation is bringing together traditional and emerging benchmarks in medical science and AI/data science. The former refers to objective scientific methodologies such as RCTs and meta studies often arranged in a top-down pyramid.

The latter often refers to relatively subjective notions such as “human-centeredness” and “trustworthiness” in the physician-patient relationship. These are beginning to be seen as essential prerequisites for Artificial Intelligence and data use in health alongside scientific evidence for safety and efficacy. 

What approaches will this PathFinder take to tackle these problems?

I-DAIR will focus initially on what kind of human-centered benchmarks are needed, how they should be developed and deployed and what could be the attributes of trusted and neutral platforms that act as “social stock exchanges” for these benchmarks.

To achieve this, we will:

  • Explore tiered approaches to benchmarks as per risk (mitigation) and agency (enhancement);
  • Consider the use of stories (micro-narratives) as a way to benchmark the impact of digital health and Artificial Intelligence solutions;
  • Develop a Validation Network, through our clinical and impact cohorts, for a distributed assessment of algorithms against local datasets and contexts. This would mirror the international round robin test system for diagnostics.

Bridging practice and research communities through micro-narratives, data flows, experience sharing and capacity-building – BCoP

What global health problems does the BCoP PF engage with?

There are four challenges that this PathFinder seeks to address:

  • Lack of opportunities for collaboration across public, private and academic sectors and the consequent fragmentation of R & D effort;
  • Absence of common language among designers, researchers, developers, deciders and practitioners; lack of shared appreciation of ethical and human rights concerns;
  • In silico prospective tech solutions without integration of cross-domain socio-economic, health and policy contexts;
  • Insufficient engagement of youth, health workers and of innovators in research and development of digital health solutions.

What approaches will this PathFinder take to tackle these problems?

Through this PathFinder, I-DAIR will explore:

  • Digital ecosystems for healthcare, health research and health promotion stitched together by diagnostic and other data flows.
  • Narrative based training for health professionals and researchers to shift their mindsets from “competitive” to “collaborative” and from “domain-specific” to “cross- domain”;
  • Youth challenges, fellowships and experience-sharing on data collection, annotation and use for low or no information settings.
  • ‘Tinkerers forge’ for collaborative development of innovations

I-DAIR will also consider supporting the work of WHO and other international organisations on capacity-building through innovative research on:

  • Micro-learning for health professionals using Artificial Intelligence;
  • The early integration of data science into undergraduate medical education and vice versa, say by taking medical imaging to undergraduate engineering classes; integration of ethical and social science perspectives into both medical and computer science education.

A pilot project on narratives to bridge health worker and patient communities, and develop a shared vocabulary, which cross-links to our benchmarking work, is under way.

In this phase, we will curate early thinking from the above approaches into a roadmap for digital health research capacity-building integrating inputs from the other PathFinders.

Governance for AI and data for health

What global health problems does the Governance PathFinder engage with?

Digital Health could be stunted by public concerns on data security, ownership, privacy & informed consent; AI use could be stymied by insufficient trust in blackbox algorithms, biased training data, and discriminatory and exploitative use of predictive analytics. At the same time, there is an opportunity to use digital technologies to improve governance of health systems, enhance transparency and accountability, and raise the salience of patient and community voices.

At a global level, the challenge is fragmented approaches to digital governance and growing science/data nationalism. There is also a messaging problem on digital health (over and under selling), which leaves leaders confused about choices and investments. In many parts of the world, there is a fundamental lack of digital governance capacity. Overall, there is a surfeit of policy principles and abstraction and limited evidence in practice for proposed governance solutions.

What approaches will this PathFinder take to tackle these problems?

Our vision of Artificial Intelligence and data for health governance is tiered, multi-stakeholder and distributed. This vision is anchored in a balanced approach to tackle misuse and missed use of data as well as missing data (‘3 Ms’).

In the first year, I-DAIR is working with Accenture Development Partners and PriceWaterhouseCoopers (PwC) to develop options for its governance structures as well as its target operating model. The idea is to capture not only the state of the art but also look over the horizon to what governance and operating modalities are apt for multilateral scientific collaboration in a post Covid-19 world.

I-DAIR’s future governance work will benefit from the learnings emerging from our PathFinders. It will be informed by follow up to international discussions, including:

  • The recommendations of the UNSG’s High-level Panel on Digital Cooperation and the Road Map for Digital Cooperation;
  • The discussions on digital health at the WHO and other competent forums;
  • The work of UNESCO on Artificial Intelligence Ethics through the Ad Hoc Working Group on Artificial Intelligence Ethics;
  • The work of the Financial Times/Lancet Joint Commission on Governing Health Futures.

The Graduate Institute’s Global Health Centre, where the I-DAIR incubation is based, will also play a role in collating and researching iterative insights on health governance coming out of the PathFinder projects.