Meet the Pretoria team using AI to spot lung disease in as little as 45 seconds
In many clinics across the world, the story often unfolds the same way. A patient arrives with a persistent cough, but due to limited resources, they may be sent home or not adequately assessed. A chest X-ray is taken, but the subtle findings that could signal early-stage lung disease have to wait weeks for a radiologist's review that may never come. Lots of cases of treatable diseases like tuberculosis (TB) go undiagnosed, not due to a lack of cures, but a lack of access to timely and accurate screening.
This is the challenge a dedicated team in Pretoria, South Africa, decided to confront. Nexus Intelligence was founded in 2022 as a joint initiative between healthcare innovators Dr. Gerhard Ferreira, a medical doctor with 30 years of experience, and Andries Vorster, a clinical engineer with 25 years in the field. They saw firsthand how diagnostic delays were failing communities and knew that technology could bridge the gap. Leveraging Google’s CXR model, their team developed Nexus AI, a CE-certified medical device designed to support clinicians in screening chest X-rays for lung abnormalities, including findings suggestive of pulmonary TB.
The tool is designed to support a clinician’s interpretation of the X-ray, right at the point of care. It uses two parallel AI models for its analysis. The first determines if the lungs appear normal and automatically highlights any abnormal finding it detects. Simultaneously, a second, specialized model screens for findings suggestive of TB. In as little as 45 seconds, the healthcare provider sees the original X-ray enhanced with heatmaps and contour lines that bring every potential finding to their attention. The clinical judgment remains firmly in their hands; the AI serves as a powerful assistant for triage and analysis.
The technology's impact is already being felt in some of the world's most challenging healthcare environments. It supports national TB programs in Ivory Coast, anchors mobile screening studies in the Northern Cape, and is being deployed across South Africa and Sierra Leone. The platform's footprint continues to expand through specialized pilots supported by the Stop TB Partnership in Cameroon and Vietnam. In a powerful demonstration of its adaptability, it was used to screen miners for occupational lung diseases like silicosis nearly four kilometers underground at South Africa's Mponeng Gold Mine.
In all of these settings, the tool can run entirely offline. This is a transformative feature, as the communities bearing the heaviest disease burdens often lack reliable internet access. As one clinician noted, "What stands out is how many patients we are now identifying who could have been missed. Without the AI flagging subtle findings, they would not have been referred for further testing."
Google helped enable this breakthrough by licensing our CXR AI model, which allowed the Nexus Intelligence team to build their clinical-grade tool without starting from scratch. The platform's effectiveness was highlighted by an independent evaluation in The Lancet Digital Health, which recognized it as a top-performing screening tool. The device has also received a CE Class IIb certification for medical use, further validating its clinical readiness.
To date, Nexus AI has processed over 25,000 X-rays across 40 locations in six nations.
Ultimately, this success belongs to the innovators in Pretoria who saw a local challenge and met it with global technology, creating a future where life-saving diagnostics are within reach.
