Computer-aided detection for TB
Chest x-rays (CXRs) are one of the most sensitive tests for detecting pulmonary TB and are used as screening and diagnostic tools. However, CXRs are limited due to scarcity of resources and trained personnel, technical limitations, high cost of hardware, and intra- and inter-reader variability.
A potential technological answer is computer-aided detection (CAD) – artificial intelligence-driven machine learning, specifically deep (learning) neural networks – that can analyse CXR images for the presence of abnormalities suggestive of pulmonary TB.
How can CAD improve TB screening?
- Help radiologists optimize their workflow
- Alert human readers to abnormal images
- Requiring prioritization
- Perform pre-reading assistance
- Provide reporting assistance
- Provide quality control
- Assist teleradiology and telemedicine service
WHO recommendation
Among individuals aged 15 years and older in populations in which TB screening is recommended, computer-aided detection software programmes may be used in place of human readers for interpreting digital chest X-rays for screening and triage for TB disease.
Resources for portable
x-ray systems and CAD
