Goal


Support fast, easy and accurate computer-aided tuberculosis detection to rapidly screen and triage risk groups in resource-constrained regions. The medical software quickly analyses a digital chest X-ray image and provides insight into TB-related abnormalities. By facilitating rapid screening and not requiring highly skilled readers, the software provides a highly-accurate and cost efficient automated TB detection.  

Medical Application

Screening for the presence of tuberculosis related abnormalities. 

Timeline

Collaborating Partners

Contribution by Thirona

Delft’s CAD4TB software is powered by Thirona’s computed image analysis. Deep Learning algorithms trained on generalizable datasets combined with Thirona’s advanced quantification capabilities, deliver robust abnormality indication with a heatmap on possible affected areas. Thirona is the full software manufacturer. 

Background

Tuberculosis (TB) is one of the deadliest infectious diseases in the world. It affects over 10 million people with an annual mortality rate of 1.5 million in 2020 . However, when diagnosed in time, TB is generally curable within 6 months. Unfortunately, about one-third of patients fail to receive a quick and accurate diagnosis or effective treatment, making them more likely to die from this curable disease.  

Solving the problem of tuberculosis requires a diagnostic solution that is affordable, fast, accurate and easy to implement in resource-constrained and hard-to-reach regions. Delft’s innovative approach to translating available technology into practical solutions for people in these regions, was the driving power behind the project. Thirona was the go-to-partner for reliable AI analysis software.   

CAD4TB is an affordable solution for early detection of TB in patients who otherwise wouldn’t have access to a costly TB-test routine, risking severe development of the disease or even death. CAD4TB was designed to help non-expert readers detect tuberculosis more accurately and cost-effectively. The software is also available in offline mode, making it suitable for very rural regions without internet, allowing for more costly TB tests to be used only in suspected cases.  

Results


CAD4TB is a class IIb medical software that instantly analyses a chest X-ray image and gives an indication of abnormalities present that can be related to pulmonary TB. The software has been CE-certified for use on ages 4+, and is available for online and offline use. It has received expert feedback from lung specialists, validating it as extremely accurate and reliable with no need for medical specialist involvement. Its performance and efficiency have also been strongly validated.   

The software has been extensively used for TB projects around the world supported by the WHO, Global Fund, USAID, Stop TB Partnership / TB REACH, EU and the World Bank. The software is integrated into the Screening, Analytics and Management platform, thus facilitating data governance and improving data access. This allows for connectivity with other data systems at a national level, and the software can serve as the first data entry point for health workers, ensuring National TB Programs’ access to aggregated data. CAD4TB has already impacted more than 14 million people, and is used in 50+ countries worldwide.  

Ready to apply AI image analysis to get better outcomes faster?
Let us help you find the perfect solution for your projects!!


GET IN TOUCH!

Share this content: