Over 100 publications
Welcome to our comprehensive collection of research publications. Here, you’ll find an extensive range of articles, scientific papers, and studies that showcase the impact of CAD4TB, portable X-rays, and more on the field.
Learn how we are shaping the future of global health together.
Highlights on CAD4TB Performance in Publications
CAD4TB had the highest overall accuracy (73.8% specificity at 90% sensitivity), was significantly more specific than other algorithms, and achieved the minimum WHO target accuracy for a TB triage test.
The AUC (95% CI) of CAD4TB against the microbiological reference standard (Xpert Ultra and/or sputum culture positivity) was 0.90 (0.82-0.97).
As threshold determination must be context-specific, our analytically-straightforward approach should be adopted to leverage prevalence surveys for CAD threshold determination in other settings with a comparable proportion of eligible but not tested participants.
The newer version (CAD4TB 7) significantly outperformed the predecessor (CAD4TB 6), performing better than human readers and meeting WHO TPP values.
CAD4TB is an accurate tool for community-based TB screening for prevalence surveys in Kenya. CAD4TB 6 met the optimal WHO TPP.
Highlights on CAD4TB Efficiency & Effectiveness in Publications
CAD has the potential to be a useful and cost-effective screening tool for TB in a resource-poor HIV-endemic African setting, assisting active case finding strategies to break the TB transmission cycle.
In the context of community-based ACF in endemic TB/HIV settings, using POC Xpert and x-ray screening with CAD analysis is both feasible and had high diagnostic yield for TB and COVID-19.
Ensuring CXR for all can be operationally challenging in a programmatic setting, however, the combined use of digital X-ray and CAD4TB in this project replaced the human processing and interpretation of X-ray and contributed to the high CXR coverage.
The proportion of people diagnosed with TB who had symptoms was very small. This suggested that most of the detected TB cases (>80%) were asymptomatic and were captured by the contribution of CXR.
Delft Light portable digital X-ray and CAD4TB in parallel with the WHO 4-symptom screen achieved a lower pre-diagnostic loss of presumptive TB cases.
The W4SS + portable X-ray with CAD screen-based intervention was more efficient for TB case yield: four times the number of TB cases yield than symptom-only-based screening intervention.
The per-screen costs for the two CAD software programs with a perpetual licensing costing structure are considerably lower than the cost with radiologists for high throughput in ACF scenarios.
With high throughput scenarios, the per-screen cost for CAD4TB is 73% lower than a radiologist for ACF and 61% lower for facility-based screening.