Page 66 - SAMRC AnnualReport 2025-26

AI in Medical and Health Research

The Unit’s research is anchored in high-quality,
community-representative population data
platforms supported by sustained public
engagement.
The Agincourt HDSS tracks births, deaths and
migration across a whole-population cohort of
approximately 120,000 people in 31 villages,
generating dynamic evidence on health transitions
used by local and national decision-makers.
The platform is linked to clinic and district hospital In partnership with the Wits Entrepreneurship
systems, enabling secure longitudinal linkage of Clinic and supported by BANKSETA, the
patient records for adults accessing healthcare EMU delivered a six-month entrepreneurship
services. This infrastructure supports: programme for 63 entrepreneurs from
Agincourt villages.
› Development of electronic health records
› Multimorbidity dashboards piloted with district
and clinic managers
› Population-level analyses of healthcare utilisation
and unmet need

The Unit is also introducing advanced technologies
into underserved settings:

› Nurses are conducting AI-assisted
echocardiography to improve detection of heart
failure associated with widespread hypertension.
› Portable low-field MRI technology, introduced in
collaboration with Columbia University, is being
evaluated alongside high-field MRI to expand
dementia research and explore future primary
healthcare applications.
› Imaging data generated through the IMCI-Plus
Trial are contributing to AI-enabled models for
point-of-care ultrasound to reduce inappropriate
antibiotic prescribing in childhood pneumonia.

› The Kuamsha app, developed through a South
Africa-UK partnership, demonstrated strong
uptake and engagement among adolescents
as part of research addressing widespread
depression. Further refinement is underway in
preparation for a full-scale trial.
Presenters at the entrepreneurial programme.

64 SA M R C A N N U A L R EP O R T 2 0 2 5 / 2 0 2 6