Case Study

Kenya – Enhancing TB case detection in community outreach using Delft Light & CAD4TB

Delft Light being used by Centre for Health Solutions in Kenya

The National TB Program of Kenya received eight Delft Light ultra-portable X-ray machines with CAD4TB through USAID’s iNTP. These were deployed in seven counties and integrated into Kenya’s TB-targeted outreach programs.

USAID-supported deployment

Kenya’s National TB Program received eight Delft Light ultra-portable X-ray systems with CAD4TB through USAID’s iNTP.

Multi-county rollout

The systems were deployed across seven counties.

Integration into outreach programs

Delft Light and CAD4TB were integrated into Kenya’s TB-targeted outreach activities.

Result

In 2022-2023,

15 ,916 people
were reached
56 %
had TB symptomps
  • People with CAD4TB scores >60 were sent for GeneXpert testing (5% or 793 individuals). An additional 5% with CAD4TB scores between 40-59 were tested based on clinician judgment.
  • TB positivity was high among those with high CADTB scores: 28% (163) had TB confirmed by GeneXpert, and nearly 30% (47) were asymptomatic.
  • TB positivity was 6% for CAD4TB scores of 40-60.
  • Before introducing CAD/AI-enabled X-rays, the TB positivity rate was low (5-10%).
Delft Light being used by Centre for Health Solutions in Kenya
Conclusion
  • Using Delft Light and CD4TB proved to be a more cost-effective approach for reducing GeneXpert tests in outreach settings.
  • These tools effectively identified individuals for further CAD4TB scores. This underscores the effectiveness of AI-enhanced CXR in early TB detection through community-based screening.
  • We recommend expanding this technology to improve TB detection in Kenya.

REFERENCE: Krisa, R. et al (2024, November 12-16). Enhancing TB detection through integrating computer-aided detection (CAD) software in chest radiography: Findings from targeted outreach programs in Kenya [Presentation]. The Union World Conference on Lung Health, Bali, Indonesia.