On March 11th 2022, 25 researchers from the VU Campus Center for AI & Health and the Amsterdam Institute for Global Health and Development (AIGHD) came together to spend an afternoon exploring research gaps around the theme of ‘AI for Global Health’.

In low- and middle income countries (LMICs), healthcare facilities are often extremely limited and understaffed. AI-based technologies have the potential to improve health outcomes by assisting overburdened and sometimes poorly trained doctors, nurses and community health workers in analysing medical data and supporting clinical decision-making. However, bringing AI-based decision support systems to low-resource settings is not without challenges. In addition to technological challenges such as dealing with a lack of data, there are important practical, legal, ethical and acceptability questions to be addressed. These various issues were the focus of the four short presentations, brainstorm sessions in small breakout groups and the plenary discussions.

(Text continues below photos)

Carolyne Egesa (medical anthropologist, AIGHD) started out by presenting some background on the reality of healthcare in low-income settings, and gave some pointers on particular healthcare challenges in these settings where AI may have a positive impact. Johnblack Kabukye (MD and informatician, AmsterdamUMC and Uganda Cancer Institute) then presented an overview of the various technological challenges of implementing AI technology in low-resource healthcare settings, using examples from his research on cervical cancer screening in Uganda. Mohammad Rezazade Mehrizi (Associate Professor of Organisational Learning, VU Campus Center for AI & Health) showed the difficulties of implementing new technologies such as AI within healthcare practice, advocating a ‘transformational’ approach: integrating an redesigning local workflows and procedures along with the introduction of the new technology. Finally, Hannah van Kolfschooten (researcher and lecturer on AI & Health Law, Law Centre for Health & Life) provided some perspectives on various legal and ethical aspects, including accountability and data privacy.

With these presentations forming the starting point for further brainstorm sessions and informal discussions, a lot of ground was covered in one afternoon. Needless to say, with such a complex and diverse topic, this still only amounts to scratching the surface. With many pertinent research gaps identified, we hope these new insights and the connections that have arisen in this theme meeting will lead to future collaborations and exciting research projects.