Medical AI company Lunit’s chest X-ray analysis software INSIGHT CXR has demonstrated superior performance in tuberculosis (TB) detection, as highlighted by an independent study.
The study, published in The Lancet Digital Health, showed that INSIGHT CXR achieved the highest accuracy out of 12 tested AI-based TB detection products. Additionally, the device showed 89.9% sensitivity at 67.7% specificity, and 89.5% sensitivity at 70.2% specificity, meeting the World Health Organization (WHO) target product profile.
Researchers at the Heidelberg University Hospital in Germany and the Stop TB Partnership at UNOPS assessed the 12 products using data from South Africa’s national TB prevalence survey. The study involved 774 participants, including 258 confirmed TB cases, and is the first to comprehensively evaluate these AI tools in a high TB setting, according to Lunit.
TB is a bacterial infection that mainly targets the lungs and spreads through airborne droplets when an infected person coughs or sneezes. The infection disproportionately affects populations with compromised immune systems, such as people with HIV/AIDS, and individuals with limited access to healthcare.
Lunit’s CEO Brandon Suh said: “Lunit INSIGHT CXR's ability to maintain high sensitivity across diverse populations and thresholds is particularly crucial in resource-limited settings, where every undetected case can have far-reaching consequences.
“By bridging the gap in TB diagnosis, we are contributing to a future where no case goes undetected, especially in regions where it matters most.”
The South Korea-based company has a range of detection software under its belt. Lunit’s INSIGHT MMG tool – which analyses mammograms to detect breast cancer – has shown a superior cancer detection rate compared to radiologists in previous studies.
The company’s suite of INSIGHT radiology solutions won CE and UK Conformity Assessed (UKCA) marking in November 2022, with INSIGHT CXR initially receiving CE marking in 2019. Both INSIGHT MMG and CXR are cleared by the US Food and Drug Administration (FDA).
Lunit secured $150m in a November 2023 capital chip-in. In a company announcement, Lunit said that about $52m of its capital proceeds will go to building new AI cancer diagnostics and therapeutics, while the rest of the funds will go towards its subsidiaries, investment, labour recruitment, and other operations.
In December 2023, Lunit signed a deal to acquire New Zealand-based Volpara Health Technologies for an A$292m ($193m) price tag. Volpara provides AI-powered image analysis, which enables radiologists to precisely quantify breast tissue. It also aids technologists in producing mammograms with optimal image quality. The acquisition was completed in May 2024.
