Scientists at Shenzhen Zhying Medical Imaging in Guangdong, China, have developed an automated detection system based on artificial intelligence (AI) for screening active TB disease on CT images.
They conducted a study to test the accuracy of this detection system that can automatically read DICOM CT images to diagnose active TB disease and also differentiate between TB and pneumonia. This AI-based system, followed with a self-organized clustering algorithm in the study, was found to accurately diagnose active TB disease during its earlier stages, and distinguish between active TB disease and non-active TB (i.e. pneumonia and normal cases). It simplifies the diagnosis process and lays a solid foundation for AI in CT diagnosis of active TB disease in a large-scale clinical application.
The sensitivity and specificity for artificial intelligence are very high (0.935 and 0.971 respectively), which show that the AI-tool performs well for diagnosis of active TB disease and differential diagnosis of active TB disease and pneumonia. It can be implemented with minimal need for human intervention. However, while presenting results of this study, Dr Fleming Lure, investigator of the study, cautioned that, "AI can function as a second pair of eyes. It does not replace the humans but provides references for healthcare professionals to make a final diagnosis. Also, this small study lays good groundwork for the future of AI in the large clinical evaluations and simplify the diagnosis process."
109 months are left to end TB worldwide: will new tools be pivotal?
The fight to end TB is a human rights imperative. We have to use all existing tools to prevent, diagnose and treat TB rationally and to maximal potential, as well as fully fund the research and development of new and better tools - and - also ensure that new tools are rolled out and reach the people in-need as soon as possible without any delay.
Shobha Shukla - CNS (Citizen News Service)
Shobha Shukla is the award-winning founding Managing Editor and Executive Director of CNS (Citizen News Service) and is a feminist, health and development justice advocate. She is a former senior physics faculty of prestigious Loreto Convent College and current Coordinator of Asia Pacific Regional Media Alliance for Health and Development (APCAT Media). Follow her on Twitter @shobha1shukla or visit www.bit.ly/ShobhaShukla
- Shared under Creative Commons (CC)
(Note: You can view every article as one long page if you sign up as an Advocate Member, or higher).
