2013
A study was published on detecting tuberculosis using automated reading (utilizing the CAD4TB artificial intelligence solution). The study should show that the assessment of chest X-rays using CAD4TB and by clinical officers was comparable and that CAD4TB had the potential as a point-of-care test and for the automated identification of subjects who require further examinations. Note that this study used a significantly older version of CAD4TB than currently available.
2014
Another study was published looking at the utilization of CAD4TB in Lusaka, Zambia. It showed that the use of CAD with digital chest X-ray has the potential to increase the use and availability of chest radiography in screening for TB where trained human resources are scarce.
2017
A study looked at using digital chest X-rays with computer-aided diagnosis software (CAD4TB) versus symptom screening to define tuberculosis among household contacts and the impact on tuberculosis diagnosis. The study showed that symptom screening if used alone with follow-on definitive TB testing, would have led to missing eight of the 19th confirmed TB cases, and CAD software could support TB screening efforts.
CAD4TB was reviewed as part of the Zambia National Tuberculosis Prevalence Survey that same year. A study on this objective looked at the performance of the CAD4TB software against that of field- and central readers. The performance of CAD4TB was similar to that of field and central readers. The study concluded that the performance of automatic chest X-ray readings is comparable to that of human experts in a TB prevalence survey setting using culture as a reference.
2021
Later, in 2021, one study was published on the costs and cost-effectiveness of a comprehensive tuberculosis case-finding strategy in Zambia, also considering CAD4TB.
That same year, a study looked at using CAD as a triage test for pulmonary tuberculosis and found that CAD-based chest X-ray analysis can be implemented as a high-sensitivity tuberculosis rule-out test.
2022
A study reviewed the performance of computer-aided detection digital chest X-ray reading technologies (CAD4TB) for the triage of active tuberculosis among persons with a history of previous tuberculosis. It showed that CAD4TB achieved a sensitivity and specificity of 89.3% and 24%, and 90.5% and 60.3% amongst those with and without previous TB, respectively.
2023
In 2023, a study assessed non-tuberculosis abnormalities on digital chest X-rays with high CAD4TB scores using data from a tuberculosis prevalence survey in Zambia and South Africa. The study found that a wide range of non-TB abnormalities can be identified on digital chest X-rays among individuals with high CAD4TB scores but do not have bacteriologically confirmed TB. A tool like CAD4TB could simultaneously identify other causes of abnormal chest X-rays alongside TB, which could be interesting for future research.
