Building infrastructure for new digital health technologies in Africa

July/August 2023 | Volume 22 Number 4

Photo courtesy of Deshen Moodley Deshen Moodley, Ph.D., co-founder of the South African National Centre for Artificial Intelligence Research (CAIR)

Deshen Moodley, an associate professor at the University of Cape Town, co-founded the South African National Centre for Artificial Intelligence Research (CAIR). After completing his master's in computer science, he worked at a series of software development companies in South Africa and abroad. Returning to the continent in 2001, he completed a Ph.D., specializing in AI, and eventually lectured at the University of KwaZulu-Natal (UKZN). He’s been a visiting researcher at Massachusetts Institute of Technology (MIT), U.S., University of Edinburgh, UK, and University of Muenster, Germany. His current research is carried out within the AI Research Unit at the University of Cape Town, where he holds the SARCHI Research Chair in AI Systems.

When did you become interested in computers?

I grew up as an Indian South African during apartheid. I was just starting high school (around 1986-87) and read somewhere about computers and was absolutely fascinated that you could program these machines. I didn’t have access to computers, which were not yet mainstream in South Africa, especially in our community, so I saved up some pocket money and bought my first computer. From there, I taught myself programming. After that, I knew exactly what I wanted to do: program computers!

How has your experience in industry influenced your career?

The massive advantage is I understand the skills our future computer science graduates will need to become employed in the field. Comprehending high-level corporate management has been incredibly valuable in terms of my own research management and leadership. Software engineering is a very professional discipline. Setting up and managing organizational structures is something I learned on the job, so I’ve assumed several leadership roles, including heading the computer science department at UKZN, based on that experience.

What brought you back to South Africa?

My last industry job was at a startup in Denver, Colorado, back in 2000. I was a typical techie and enjoyed the startup culture—despite working long hours. As I entered my late 20s, I decided that this vibrant corporate environment wasn’t enough to drive me for the rest of my life and I also felt an emotional attachment to the community that had supported me. I decided to return to South Africa, to UKZN in my home province to begin my Ph.D. and to take up a position as a lecturer. There, I was able to cultivate my own skills while developing young people and sharing my experience with them. I took a long time to complete my education because I put a lot of time into training students and teaching Java programming to doctors and health care professionals (as part of an NIH program). For me, impact was more important than building an academic profile.

How did studying abroad help you?

My first experience studying abroad was via an Informatics Training for Global Health (ITGH) scholarship at MIT in Peter Szololvits’ clinical decision-making lab in 2004-2005. It was, of course, MIT, so I experienced computer science and AI at the highest level. This completely different setting opened my eyes as to how things could be done. This experience had a profound impact on my view of building a computer science group that also studied health—a multidisciplinary group led by computer science. My other host, Dr. Hamish Fraser, who was based at Partners in Health and Harvard Medical School, worked closely with MIT and involved me in a number of active projects. We published a highly cited paper on electronic medical records in developing countries.

During the ITGH program I met Chris Seebregts from South Africa’s Medical Research Council who’s become a leader in digital health. Together, as part of the ITGH program, we helped set the curriculum, recruited students and taught courses. Having forged a partnership, we later obtained grants from the Rockefeller Foundation and the Canadian International Development Research Centre to build the Health Architecture Lab, a South African applied computer science lab for innovation in health informatics. We trained others, produced technologies, established a core body of knowledge for building national health information systems in Africa, and established a network of research collaborators. We created a massive organization and a community. After a few years, we folded the lab into a broader national center, CAIR.

What role does CAIR play in South Africa?

Tommie Meyer (a researcher I met at an ITGH event) and I primarily set up CAIR because we recognized—and this was in 2010—that AI would be an important technology going forward. CAIR was also needed because computer science departments in South Africa were very small then. Usually, they had between 15 and 20 faculty members—and that would be a large IT department. These numbers cannot create critical mass, especially in an important field, so we knew we had to pool resources and construct a nationally distributed research environment with the help of South Africa’s Department of Science and Innovation. CAIR is now the flagship program for building capacity for AI research in South Africa.

How does your work in South Africa influence global health research?

Along with CAIR, I work with Jembi Health Systems, an NGO which designs and builds information systems for tackling health problems. Jembi worked in Rwanda, Mozambique, and a few other continental African nations, so we produced a body of knowledge for creating national health information systems. While the original focus was on African countries, the knowledge and skills we’ve produced fit health informatics in all countries.

Settings with less access to technology constantly import it from other countries, yet there are many differences between high-income and low-income countries—resourcing, culture, political environments, and other dissimilarities. When people have tried to use non-local technology, there have been some very costly failures. I believe that it is very important to train innovators and build infrastructure for creating new digital health technologies in Africa. In this way, we can contribute to other nations in the region and beyond… and even expand to nations with much greater access to resources. My international colleagues will say, “Hang on, your ideas are equally applicable to high resource settings. Your technologies would work in our rural areas or for certain populations.”

What’s ahead in the health informatics field?

I'm working on a low-cost wearables project with my Ph.D. students. We want to dynamically use machine learning to build a model of an individual’s personal health in real-time using wearables, and then we want to integrate this monitoring and sensing into their daily activities, so they can adapt and learn their own optimal health state—we want to promote proactive techniques for health. A recent development in the field is detection of emotional well-being from an ECG signal (a heart rate signal) using a low-cost wearable. Mood is very important yet its impact on well-being is something we don't fully understand. I think something will emanate from our work very shortly.

My bigger vision is that AI and the broader advanced technologies, like wearables, virtual reality and digital twins (virtual models that accurately reflect individual bodies), could drive more prevention-based public health systems, especially epidemic control mechanisms, in African countries. My colleagues and I are going to publish some interesting pieces about that shortly. Generally, technology is under-utilized in health so there’s an opportunity for us to reimagine the ways we see medical solutions. In a low-resource setting, you cannot afford to put all your eggs into the curative basket—it's too costly and we don't have the facilities and staff. For me, prevention is the most compelling idea and I believe that AI can bring radical change to our approach to public health.