Abstract
Chest radiography (CXR) plays a key role in the detection of pulmonary tuberculosis (TB). While current World Health Organization (WHO) guidelines recommend diagnosis on the basis of bacteriological or molecular findings, the integration of CXR into screening and triaging algorithms can assist with early detection of TB, as CXR is a highly sensitive tool for detecting TB disease, and can guide the effective use of diagnostic testing to improve case detection and cost-efficiency.
Computer-aided detection (CAD) products use artificial intelligence (AI) to analyse CXR images for the presence of abnormalities suggestive of pulmonary TB, producing an abnormality score that can be used to determine the need for follow-on diagnostic testing for TB relative to a selected threshold. CAD technology can improve the feasibility and performance of CXR for screening and triage for TB disease and may benefit TB programmes by enhancing capacity for TB screening. Such technology can replace or augment human expert interpretation of plain CXR when screening for TB and can avoid inter-reader variability and reduce delays in reading radiographs when skilled personnel are scarce.
Following a review of the evidence, a new recommendation that supports the use of CAD for adults over the age of 15 years in place of human readers for interpretation of digital chest radiography in both screening and triage for pulmonary TB disease was included in the updated WHO TB screening guidelines and recommendations. This new recommendation is expected to increase the consideration of CAD by NTPs and other partners as a tool to enhance the use of CXR in screening and triage contexts.
Effectively integrating CAD products into routine programming requires determining an appropriate CAD threshold that will be used to signal probable TB cases and trigger further TB diagnostic evaluation. Identifying the ideal threshold for each implementation requires calibration of CAD products based on the local context and intended use case, as well as decision making around the goals for screening and acceptable costs.
This toolkit has been developed by the Special Programme for Research and Training in Tropical Diseases (TDR) and the Global TB Programme (GTB) at WHO, in collaboration with partners and experts working on CAD and TB screening. This toolkit aims to support the implementation of CAD for TB screening and detection and is intended for use by national TB programmes (NTPs) and other implementers who have decided to use CAD.