Regional hubs catalyze AI to improve women’s health

Artificial intelligence (AI) and its rapidly expanding applications are transforming how health systems are planned and how health services are delivered globally, including across low- and middle-income countries.   

For example, AI can leverage data to reveal patterns in disease and improve treatment and care, predict a patient’s health risks and improve efficiencies in healthcare services and systems, among other innovations. 

AI presents significant opportunities and risks for areas of health systems that are particularly lagging, such as maternal, sexual and reproductive health (MSRH). Maternal mortality levels are not on pace to meet the Sustainable Development Goal (SDG) target, the prevalence of child marriage and teen pregnancy remains high and significant gaps remain in access to contraception, safe abortion and post-abortion care and comprehensive sexual and maternal health education.  

There is an urgent need for evidence on innovative solutions that improve MSRH and women’s rights while strengthening health systems for integrated and sustainable health services. These help women, their families and their communities to live healthy lives and thrive. 

IDRC’s Artificial Intelligence for Global Health (AI4GH) initiative is working at the intersection of this urgent need for evidence-based solutions and the capacity of AI to spark transformative positive change.  

The five-year, CAD15.5 million AI4GH initiative is funding researchers in low- and middle-income countries to strengthen health systems by leveraging contextualized responsible-AI solutions to improve MSRH for women and girls, and to support more effective and equitable preparedness and responses to epidemics and pandemics. 

The AI4GH initiative promotes responsible AI, which is to say solutions that are ethical, inclusive, rights-respecting and sustainable. It supports robust, innovative and interdisciplinary research projects and processes across sub-Saharan Africa, Asia, Latin America and the Caribbean, and the Middle East and North Africa.  

AI4GH has established four regional hubs to support the capacity of regional and local institutions to oversee research and build research networks.  

“These hubs identify promising AI research topics and teams, and provide them with grants, technical support, and opportunities to collaborate and share knowledge,” said Chaitali Sinha, a senior program specialist with IDRC’s Global Health program. “They all share a common point of departure but are fine-tuned to the local contexts they represent. This is part of the AI4GH approach.” 

Beyond leading the AI4GH work in each region, the hubs will collaborate to contribute to cross-regional learning and knowledge production. They will also identify, validate and document priority themes, strengthen the capacity of innovators, researchers and policymakers, and strengthen the number of homegrown AI innovations in MSRH and rights. 

AI health solutions include predictive analysis, mobile tools, telemedicine platforms and diagnostics to improve the accuracy of illnesses, particularly in contexts where specialized medical staff are unavailable. 

In their efforts to develop innovative healthcare solutions, the hubs will also mitigate risks such as privacy, according to Samuel Oji Oti, a senior program specialist with IDRC’s Global Health program.  

“Privacy and confidentiality considerations relating to the use of AI interventions are critical to protecting vulnerable women and girls from unintended consequences such as stigma, discrimination and intimate partner violence,” he said. 

Each of the four AI4GH regional hubs is designed to advance MSRH and rights while strengthening health systems within their region through the responsible development of AI innovations.  

Progress has been particularly slow in Africa toward universal access to sexual and reproductive health and rights. Yet, an explosion in available health data in Africa due to improved infrastructure for electricity and internet, as well as the widespread adoption of digital health technologies, provides great opportunities for data-driven innovations. 

The Africa hub — known as the Hub for Artificial Intelligence in Maternal, Sexual and Reproductive Health in Africa (HASH) — so far has more than 100 members. 

Ten sub-grantees have been chosen to implement AI projects on four priority themes: maternal health, sexually transmitted infections, adolescent sexual reproductive health and HIV. The projects are being implemented in Nigeria, Ethiopia, Namibia, Tanzania, Ghana, Kenya and Uganda. 

The hub will offer a platform for collaboration, innovation and learning, according to the hub’s co-principal investigator, Dr. Rosalind Parkes-Ratanshi, who is the founding director of the Academy for Health Innovation, based at the Infectious Diseases Institute at Uganda’s Makerere University. 

“There are many ways AI tools can be used to improve access to care, enable early interventions and improve patient outcomes in Africa,” Parkes-Ratanshi said. 

Parkes-Ratanshi adds that HASH’s hub will work to address challenges facing the region, including a low understanding of AI’s potential and risks, a lack of clarity on ethics and research regulatory procedures and a shortage of digitized data tailored to the African context that fuels data bias. 

In lower-income countries across Latin America and the Caribbean (LAC), indicators for MSRH are not on track to meet SDG targets. The challenges are particularly pronounced for vulnerable groups, such as women, adolescents, Indigenous populations, rural populations, Afro-descendants, migrants, people with disabilities and LGBITQ+ people, among others.  

Of the 10 million women who give birth each year across the region, 8 percent of them do not have access to adequate prenatal care and 7,600 die from causes related to pregnancy and childbirth. The Caribbean is one of the regions of the world most affected by the HIV epidemic, with 23 percent of people living with HIV not knowing their diagnosis. 

The innovative use of AI solutions presents the potential to improve outcomes on these and other longstanding health issues. Leading this work in the LAC region is Argentina’s Center for Implementation and Innovation in Health Policies (CIIPS-IECS). 

“We found that LAC is one of the least-developed regions in terms of countries with published artificial intelligence strategies,” said Cintia Cejas, the hub’s principal investigator and a coordinator at CIIPS-IECS and the Center for Artificial Intelligence and Health for Latin America and the Caribbean (CLIAS). “The maturity of AI in health in our region could be considered exploratory in some countries and emerging in others, but none has evidence of a truly integrated AI ecosystem.” 

The LAC regional hub will support an implementation research network on responsible AI innovations to advance MSRH and rights across the region. The hub will provide financial and technical support to selected AI researchers and innovators to strengthen research quality, integrate social and gender analysis, and position the results to achieve influence and optimal scale. 

South Asia faces multiple challenges that include high rates of unplanned pregnancies, maternal deaths, unsafe abortions, child marriages and sexual violence. 

The challenges are well-known to Mohammad Imran, executive director of PHC Global, which is leading the South Asia hub, known as AI-SAROSH, along with the GTA Foundation.  

One illustration of the challenges is the region’s rate of 157 deaths per 100,000 births, well behind the SDG target of 70 deaths per 100,000 births. 

“An estimated 78,000 died due to reasons surrounding pregnancy and childbirth in 2017 — many of which could have been prevented with proper antenatal, obstetric and perinatal care,” Imran said.  

AI tools hold enormous promise for improving public decision-making, increasing productivity and overcoming human resource constraints in public health.  

“We aim to create a regional hub to catalyze multi-disciplinary approaches in solving key issues in [MSRH] in South Asia through AI technologies and bridge gaps between various disciplines,” Imran said. “We will prioritize AI innovations that target the population, individuals, health systems and supply chain levels of [MSRH] in the region.”  

Protracted conflicts have further fragmented health systems, triggered significant population displacement and sparked migration, explains the representative of the Middle East and North Africa hub, Nour El Arnaout, who is a division manager at the Global Health Institute at the American University of Beirut.  

“Many countries suffer from relatively high rates of maternal mortality and morbidity — haemorrhage, gestational diabetes, hyperglycaemia, among others — that remain the leading concerns threatening pregnant women,” she said. 

The region has high rates of early marriage, unsafe abortions and gender-based violence, particularly among adolescents, refugees and displaced populations. 

Further barriers to health outcomes include a lack of trained health providers, a lack of awareness among women about the importance of accessing MSRH services, poor quality of these services, economic barriers, and protracted conflicts and crises. 

This hub is working to leverage opportunities presented by AI, including AI-based mobile applications targeting health providers, AI-based prediction and triage models or algorithms that could flag complications in need of medical assistance, AI-enabled chatbots and AI-automated messages to selected target populations. 

Together, the four hubs will play a crucial role in exploring how research can be leveraged to catalyze the responsible development and deployment of AI innovations for advancing MSRH and rights as part of strengthened health systems across low- and middle-income countries.