In a significant leap forward in the battle against tuberculosis (TB), Stellenbosch University (SU) researchers are spearheading a pivotal global trial aimed at revolutionising TB diagnosis through the innovative use of artificial intelligence (AI). The project seeks to create and validate an algorithm designed to empower healthcare workers at primary care facilities to identify potential TB cases swiftly and accurately, utilising a handheld ultrasound device that connects to a smartphone.

“TB remains the world’s deadliest infectious disease, yet it is massively underdiagnosed,” explained Prof Grant Theron, a leading expert in Clinical Mycobacteriology and Epidemiology at SU and the trial coordinator. He said the critical challenges faced in diagnosing TB, where many patients undergo unnecessary testing, while others who desperately require screening are overlooked. “There’s an urgent need for accessible, affordable, and scalable diagnostic tools for TB triage,” he asserts.

This ambitious project, titled ‘Computer assisted diagnosis with lung ultrasound for community based pulmonary tuberculosis triage in Benin, Mali, and South Africa’ (CAD LUS4TB), involves an esteemed consortium of 10 health and research institutions across Africa and Europe, with financial backing of €10 million (over R200 million) from the European Union’s Global Health EDCTP3 Joint Undertakings.

The study plans to enrol 3,000 adult patients, aiming to assess the efficacy of AI-driven ultrasound technology in enhancing TB detection and management. The overarching goal is to significantly increase access to TB screening for symptomatic adult patients at the primary healthcare level, ensuring timely intervention where it is needed most.

“Point-of-care lung ultrasound employs sensitive, handheld imaging devices capable of detecting body abnormalities, including those characteristic of TB,” Theron said. Historically, this technology has suffered from a dependency on specialised expertise for image analysis. However, thanks to advancements in AI, there is a newfound opportunity to automate image interpretation, enabling even minimally trained health workers to swiftly discern which patients require additional testing. The CAD LUS4TB initiative thus heralds a transformative, specimen-free diagnostic solution in the ongoing fight against TB.

In collaboration with European partners, SU will delve into developing and validating cutting-edge machine learning algorithms, leveraging the expertise of Prof Thomas Niesler’s Digital Signal Processing group within SU’s Faculty of Engineering. The goal is to craft a sophisticated algorithm compatible with portable ultrasound devices linked to smartphones. This innovation promises to provide automatic assessments of ultrasound images for TB indicators, all encapsulated within a user-friendly mobile application ready for widespread use.

The project is set to commence on September 1, 2025, under the dynamic co-leadership of Dr Veronique Suttels from The Swiss Federal Technology Institute of Lausanne and Prof Ablo Prudence Wachinou from the National Teaching Centre for Pneumology & Tuberculosis in Benin. Together, they aim to lay the groundwork for a brighter future in TB diagnostics.

The CAD LUS4TB consortium is committed to generating robust, population-specific evidence while advocating for the integration of computer-assisted diagnosis (CAD) powered by AI, thereby influencing healthcare policies related to lung ultrasound implementation.

Source link