New unit a shot in the arm for healthcare research in Africa

The newly established South African Medical Research Council (SAMRC) and University of Cape Town (UCT) AI for African Population Health Unit is set to create new opportunities for collaboration between researchers and clinicians, harnessing diverse health data to improve healthcare outcomes across the continent.

Leading the new unit is Professor Nicola Mulder, who brings her expertise in data-driven science and capacity development. Located at the Computational Biology Division at the Faculty of Health Sciences (FHS), the unit started operating this month.

The SAMRC Extramural Research Units (EMUs) are external divisions commissioned to conduct strategic health research. Of the seven new EMUs officially launched on 28 July, two are based at UCT: the SAMRC /UCT AI for African Population Health Unit; and the SAMRC / UCT Discovery Neuroscience in Children: Advancing Understanding and Treatment of Acute Brain Conditions Unit.

From left: Prof Ntobeko Ntusi, president and CEO, SAMRC; Prof Nicola Mulder, head of the new SAMRC /UCT AI for African Population Health Unit; Prof Tyrone Pretorius, chairperson of the SAMRC; and Prof Liesl Zühlke, vice-president of Extramural Research and International Portfolio, SAMRC.Photo SAMRC.

The SAMRC now has 30 EMUs, which see research being done by different South African and international science councils, medical schools, universities and research institutes. Explaining the objective of the new unit, Professor Mulder said: "The main aim of this proposal is to develop, test and apply artificial intelligence (AI) / machine learning (ML) tools tailored to African population health contexts and create a framework for the integration of multimodal data and the application of AI to address local health challenges.

"The tools will be trained on comprehensive African biomedical and clinical datasets to enhance the early detection of prevalent non-communicable diseases (NCDs) in South Africa (such as cancer, diabetes and cardiovascular diseases) and facilitate improved diagnosis, risk assessment and clinical management of infectious diseases (such as tuberculosis, HIV and malaria). The initiative will use state-of-the-art AI and ML techniques to facilitate the discovery of precise biomarkers, improve diagnostic accuracy, and support personalised therapeutic choices for diverse African patients in various healthcare settings."

Using African data

She said the unit's work would involve using different types of African data, together with AI techniques, to uncover the mechanisms of diseases and the interaction between non-communicable and infectious diseases. This information would be used for better diagnosis, treatment and disease risk prediction.

"It will train postgraduate students to develop interdisciplinary skills, with a focus on previously disadvantaged students. It will also look at the ethics of AI to ensure responsible and ethical use."

The unit includes two key mid-career scientists, Dr Musalula Sinkala and Dr Hocine Bendou, whose research involves the application of data science to biomedical research.

"They are making major contributions to developing capacity in data science. The unit will also include collaborators at UCT with expertise in biomedical imaging, proteomics and ethics and, of course, postgraduate students," said Mulder.

UCT expertise

Speaking about the unit being based at UCT, Mulder said the university has good expertise in this area. It is well connected with clinicians, biomedical researchers and core facilities in the FHS and the Institute of Infectious Disease and Molecular Medicine (IDM).

"This enables the work to be grounded in ongoing cutting-edge research and data generation. Our division also has access to significant computing facilities at UCT required for this research."

The FHS and IDM are particularly notable in their excellent science and publication outputs.

Mulder has a PhD in Medical Microbiology with a focus on molecular biology. She developed bioinformatics skills while working at the European Bioinformatics Institute; and has been leading pan-African initiatives developing capacity in bioinformatics for more than a decade at UCT.

"The field of AI for health is progressing rapidly and we are excited to be part of and contributing to it in Africa."

"For the last four years I have been running an open data science platform for an African Data Science consortium, DS-I Africa, to enable the application of data science techniques to address African health challenges," she said.

At the new SAMRC/UCT AI for African Population Health Unit, Mulder will support mid-career academics, ensure high-quality multidisciplinary science, encourage collaborations and promote the outputs of the unit.

"I will also work with unit members to develop a training programme for postgraduate students," she explained. She feels excited about this new challenge. "The unit provides funds to support new students and helps to coalesce the work of different academics and students around a common theme. The field of AI for health is progressing rapidly and we are excited to be part of and contributing to it in Africa."

Mulder still has several other projects on the go at UCT at various stages of completion. "I will work with the individual project managers on those projects to ensure they deliver. I will also continue to supervise postgraduate students."

On how her work has brought her to the field of AI, she said: "Bioinformaticians have been using machine learning techniques for a long time, so expanding this work to other AI techniques is a natural progression of the work we do. As the datasets we work with have become bigger and more heterogenous, so has the need for embracing AI increased."

Harnessing the power of AI

Professor Liesl Zühlke, vice-president of Extramural Research and Internal Portfolio at the SAMRC, said the unit represents an exciting step forward in harnessing the power of artificial intelligence to improve health outcomes across South Africa and Africa.

"The SAMRC is proud to be associated with the AI for African Population Health Unit and warmly welcomes it to the SAMRC family. We look forward to supporting its important work."

Leaving a legacy

The legacy she would like to leave at UCT will see a new generation of multidisciplinary researchers, including mid-career academics - hopefully progressing to senior academic leaders - and postgraduate students; as well as advancement in the use of AI to solve African health challenges, and research outputs that have a positive impact on the field.

Mulder said she is passionate about capacity development: "[I love] seeing young, enthusiastic students inspired by their work and developing into internationally competitive researchers."

She added that if she could describe her field of study to a classroom of primary school children, she would tell them: "We take biological data, pass it through some computational tools and produce knowledge to better understand how and why we get diseases and what to do about it."

The mood in the new AI for African Population Health Unit is upbeat. "The team is excited by this opportunity, which highlights the research in the group on AI for African population health and enables us to support postgraduate students and develop new collaborations," said Mulder.

View the full list of the seven new units.