Register for free to
view the full innovation profile.
Computer Aided Detection for Tuberculosis is an artificial intelligence system that uses deep learning techniques to automatically detect abnormalities indicative for Tuberculosis.
About
The key innovation is in the software: with today’s digital X-ray systems a high-quality chest radiograph is presented on a computer screen within a few seconds. This created new opportunities to detect TB cases faster and at lower cost. With the support of research institutions, universities and the WHO/Stop TB Partnership we have been investing in the development and CE certification of our Computer Aided Detection software for TB (CAD4TB) over the last 8 years. With excellent results: CAD4TB using deep learning to automatically analyse a digital chest image in less than half a minute and scores it between 0 - 100 on abnormalities consistent with TB. It now surpasses the performance of a trained human reader. Additionally, a ‘heat map’ of the lungs indicates the abnormal areas. With CAD4TB the teams can easily select those patients that have the highest likelihood of having TB.
Key Benefits
Tuberculosis is a deadly and infectious disease. If discovered in time, it can be cured within 6 months. The problem lies in detecting and diagnosing Tuberculosis due to lack of resources. Using artificial intelligence we have developed CAD4TB to assist governments and NGOs in lowering the prevalence of Tuberculosis in high burden countries. CAD4TB overcomes the need of an expert human reader, by indicating the likelihood of TB using a score between 0 and 100. The higher the score, the more likely the person has TB.
Applications
Using artificial intelligence we have developed CAD4TB to assist governments and NGOs in lowering the prevalence of Tuberculosis in high burden countries.