Nearly four kilometers beneath the surface of the earth, inside South Africa’s Mponeng Gold Mine, where broadband connectivity is unreliable and working conditions are demanding, an artificial intelligence (AI)system recently accomplished something few medical technologies had managed before.
Within 45 seconds, it analyzed a miner’s chest X-ray for signs of tuberculosis and occupational lung diseases, including silicosis, without relying on cloud computing or an on-site radiologist.
“Mponeng demonstrated that offline capability is real in the single most hostile environment we could find,” Dr Gerhard Ferreira, who co-founded Pretoria-based health technology company, Nexus Intelligence, alongside clinical engineer, Andries Vorster, says to FORBES AFRICA.
“That environment stress-tested the complete operating model and showed that a regulated AI medical device can function at the edge, close to the worker, in a setting with constrained connectivity, complex logistics and a high need for occupational lung surveillance.”
Loading…
Ferreira and Vorster note that one of global healthcare’s most persistent bottlenecks is not a shortage of imaging equipment, but of specialists capable of interpreting the images.
“Chest X-ray is the most-performed radiology examination in the world, and our route to scale is to integrate Nexus AI into the workflows where X-rays are already being generated, and to take AI assistance to facilities and communities with little access, too few radiologists or growing backlogs,” Ferreira says.
To address this gap, Nexus licensed a foundational chest X-ray model developed by Google Research and built Nexus AI CXR, a regulated clinical decision-support system that screens chest X-rays for tuberculosis and other significant lung abnormalities.
Rather than treat Africa as a market for imported technologies, an emerging generation of entrepreneurs, researchers, physicians and engineers are adapting AI to overcome challenges rooted in local realities.
Healthcare has, in fact, become one of AI’s most urgent testing grounds.
Africa bears roughly one-quarter of the world’s disease burden but has only about 3% of the global health workforce, according to the World Health Organization (WHO).
Radiologists remain among the continent’s scarcest medical specialists. In several low- and middle-income countries, fewer than one radiologist may serve every 100,000 people, leaving hospitals overwhelmed by imaging backlogs and delaying diagnoses for diseases where time can determine survival.
Nexus’ platform has now been deployed at approximately 40 healthcare sites across six countries, including South Africa, Vietnam and Côte d’Ivoire, where it has processed more than 25,000 chest X-rays.
The company partners with ministries of health, Global Fund-supported implementers, mobile and portable X-ray providers, mines and occupational-health programs, hospitals, radiology networks, equipment manufacturers and local resellers.
Rather than replace radiologists, the software is designed to prioritize their work by rapidly identifying likely normal studies while flagging potentially abnormal cases that would require a specialist’s attention. The result, the company says, is shorter reporting backlogs and faster patient triage in environments where radiologists remain scarce.
That distinction has become central to the company’s message.
“The shortage of doctors and radiologists across Africa is precisely why Al has the potential to transform healthcare, but adoption will only happen if clinicians trust the technology and understand the boundaries of responsibility,’’ Joseph Pategou, a biopharmaceutical professional, says to FORBES AFRICA.
He sees Al as an intelligence layer that supports healthcare professionals by organizing medical knowledge, identifying patterns, prioritizing cases, and surfacing relevant insights.
“In this model, Al supports the clinical workflow, but the physician remains at the center of care, with responsibility for diagnosis, treatment decisions, and patient outcomes.’’
Much of the global conversation around AI assumes dependable internet connectivity. Healthcare, often, cannot.
Nexus, therefore, built its platform to operate either through cloud deployment or entirely offline using dedicated, on-premises workstations. The underground deployment at Mponeng became an unusually-rigorous test.
In that type of environment, healthcare workflows cannot depend on continuous broadband access or remote computing. Instead, the AI accompanies the X-ray equipment, analyzes images locally and provides results during the screening encounter itself. Images, results and audit logs remain stored on the device before synchronizing later when connectivity becomes available.
For occupational health programs, the operational implications extend beyond technology.
Screening workers at the point of care reduces unnecessary travel and minimizes production disruptions, among other factors. What appears to be a technical feature in a hospital becomes an operational necessity underground.
Medical AI has attracted notable investment globally, but regulators emphasize that accuracy alone is insufficient.
Nexus spent several years navigating regulatory approval, moving from its Google Research licensing agreement in 2023 to obtaining CE MDR Class IIb certification under the European Union’s Medical Device Regulation in late 2025.
Also, offline deployment does not mean uncontrolled deployment. Each installation operates using version-controlled software releases, maintains local audit trails, records inference results and timestamps, and supports secure updates during scheduled maintenance windows.
Clinical quality assurance, post-market surveillance and human oversight remain part of routine operation. Users are trained on the software’s intended role, while abnormal findings follow established referral pathways.
“The clinical decision remains with the clinician. Patients never interact directly with the model. We wanted a partner committed to understanding that safety in this field was of the utmost importance,” says Lorna Omondi, Strategic Partnerships Lead at Google Research Africa.
Google Research originally developed its chest X-ray model for research into tuberculosis risk detection and lung abnormality triage.
Rather than commercializing the technology itself, Google licensed the foundational model to Nexus, whose founders adapted it into a regulated medical device for clinical deployment.
Google has expanded its work into areas including food security, African language technologies and community-focused AI research.
“Our thesis is that research delivers impact most effectively when it is placed in the hands of people with deep clinical and community context,” Omondi says.
She points to three principles that shaped the partnership: clinician oversight, equitable access and responsible data governance.
Patient data remains under the control of Nexus and its healthcare partners, she says, and is not returned to Google. Meanwhile, the partnership includes a commitment to provide more than 10,000 in-kind screenings annually for underserved communities.
“A tool that depends on reliable internet would exclude precisely the communities this work exists to serve.”
The company is attempting to scale not through direct-to-consumer healthcare, but by integrating into existing public health systems.
Organizations can purchase screening on a pay-per-scan basis, through prepaid screening bundles or
“The commercial principle is affordability at scale. As screening volume increases, the effective cost per examination declines,’’ Vorster says to FORBES AFRICA.
Nexus has positioned itself within existing global procurement systems. And, although the current regulated platform focuses primarily on tuberculosis and normal-versus-abnormal chest X-ray interpretation, Nexus is expanding into broader lung health.
Beta versions already include models for occupational diseases such as silicosis and pneumoconiosis.
The longer-term strategy is to evolve beyond a single algorithm into an integrated lung-health ecosystem that combines AI screening with clinical management software and centralized radiology reporting.
Femi Oke, a Nigerian AI product manager specializing in healthcare interoperability, argues that the continent’s greatest challenge is less about algorithms and more about trusted data systems.
“Trust must be designed into the workflow, not added after the model is trained. In healthcare, open standards such as HL7 FHIR can provide a common exchange layer, while consent, provenance and data-quality controls ensure that local data is treated as governed infrastructure, not free raw material,” Oke shares with FORBES AFRICA.
He argues for offline-first systems built around open standards, transparent recommendations, clinician accountability and locally-governed data.
This philosophy extends beyond healthcare.
Nearly 60% of sub-Saharan Africa’s workforce depends on agriculture, according to the World Bank. Yet, the overwhelming majority of farmers cultivate small plots vulnerable to climate shocks, unpredictable rainfall, pests, poor market access and limited extension services.
Dr Samuel Babatunde, Director of Operations at Nigerian agricultural production company, SBZ Development, believes similar AI approaches could transform African agriculture through precision farming, weather forecasting and improved re
“Evidence indicates that AI is able to increase crop production by about 15%–30%, lower input costs by 10%–25% with more efficient use of reh weather prediction, soil analytics, and digital market intelligence,’’ he tells FORBES AFRICA
For both healthcare and agriculture, he argues that the opportunity lies not in, simply, adopting AI but in developing affordable solutions tailored to local realities.
“Besides increasing production, AI’s biggest contribution is reducing uncertainty by predicting pest outbreaks, spotting nutrient deficiencies early, forecasting climate risks, and linking farmers to better markets to bolster resilience and food security,” he explains.
“As a professional at the nexus of environmental sustainability, data-driven decision-making, and agricultural systems, I believe Africa’s greatest opportunity is not just to adopt AI but to create affordable, locally relevant AI solutions designed for the realities of smallholder farmers.”
According to Babatunde, AI, when integrated with climate-smart agriculture, effective extension services, and accessible digital infrastructure, has the potential to fundamentally transform smallholder farming into a more resilient, productive, profitable, and sustainable enterprise.
Back in Pretoria, however, the focus remains on a more immediate problem. The world is unlikely to produce enough radiologists to meet growing demand any time soon.
Experts say if AI can safely help specialists review more images, identify disease earlier and extend diagnostic services into communities where radiologists rarely practice, its greatest contribution may not be replacing physicians at all.
“More than 90 healthcare workers use it every day, and every one of those interactions ends with a clinician making an informed decision for patients,” says Omondi.
“That is what helpful AI looks like in practice: local builders solving local and global problems… We are proud to see innovative solutions make real world impact.’’
Related Topics:#Africa, #Featured, #News Letter, #newsletter, #Nigeria, #South Africa.
