Delft supports Uganda’s combat against TB

A high TB-HIV burden country, of the 90 thousand TB patients in Uganda, 12 thousand are children. Additionally, about a quarter of the TB cases are presumed to be missing.

Keeping this in mind, the National Tuberculosis and Leprosy Programme (NTLP) focussed on the use of portable digital X-rays with AI software for TB screening. They can help provide diagnostic facilities, not only in high-TB burden districts but also for community outreach.

In 2018, 5 Delft Light backpack portable X-ray systems, with CAD4TB software, were procured by the IDA Foundation. Later, in 2020, 2 more Delft Lighter with CAD4TB software was procured by the DAHW and used in collaboration with the Uganda Catholic Medical Bureau. Witnessing the impact on the field, 10 more Delft Light with CAD4TB were procured. While five were acquired under the introducing New Tools Project (iNTP) by the StopTB Partnership, the other five were provided through the Partnership for Supply Chain Management (PFSCM).

Besides portable X-rays systems and CAD4TB software, the NTLP also secured 2 mobile TB screening clinics from Delft Imaging in 2021. Facilitated through the UNOPS, each clinic featured an EasyDR multi-functional X-ray and the CAD4TB software. In the latter half of the year, they secured another 3 more mobile TB screening clinics.

In total, the installed base of Delft Imaging solutions in Nigeria covers 17 portable X-ray systems, 5 mobile OneStopTB clinics (with stationary EasyDR X-ray systems) and 23 CAD4TB installations.

In order to ensure an effective service and support infrastructure in Uganda, Delft Imaging works closely with its local in-country partner for rapid onsite support whenever needed.

Making a difference

Delft Imaging, in collaboration with its in-country partner, supports all these installations and services through long-term service contracts. Additionally, we are proudly supporting the Infectious Disease Institute of Makerere University (IDI). Since late 2021, IDI, with the help of the StopTB Partnership’s TB REACH program, has been utilising CAD4TB to investigate the performance of A.I in the rural towns of Uganda.

The CAD4TB installed in the tertiary hospitals of Gulu, Lira and Kitgum served a combined population of more than one million. Each person with presumed TB in the hospital received a digital chest X-ray that is read by CAD4TB, trained human readers and a highly sensitive molecular diagnostic test (Xpert). These outcomes were recorded and compared to that of the molecular test in order to analyse its accuracy and optimal settings. To read more about the project, please see their case study.

In the words of Michael Odit, a Project Radiographer at IDI, “AI has helped us a lot. If you have many patients, you only need to send the images to the AI box. You can work on many patients in the shortest possible time.”

Over the years, various partners have published about the impact of different X-ray systems and CAD solutions for TB screening efforts.

The Delft Light portable backpack X-rays with CAD4TB were used for the TB screenings at health facilities in high TB burden districts and for community outreach. Between June 2020 and May 2021, more than a thousand individuals were X-rayed in health facility settings. Of those screened, 21% HAD ABNORMAL X-ray, 86% were tested by GeneXpert and the remaining 35% were confirmed with TB. In community settings in March 2021, more than 1400 individuals were X-rayed. Of whom, 13.5% had abnormal X-rays and were tested by GeneXpert. About 6.3% were confirmed with TB.

It was concluded that TB screening by mobile digital X-ray in health facilities had a higher yield of TB when used at health facilities as opposed to community settings, particularly among PLHIV and TB contact. Thus, to improve early case detection, the national TB programs and partners were recommended to scale up the usage of digital X-ray systems with CAD4TB in the screenings of high-risk populations. In the 2022 Union World Conference on Lung Health, the stakeholders in Uganda highlighted the use of dCXR and CAD4TB for systematic TB screening among PLHIV. They concluded that this approach led to twice as high TB yield among PLHIV and contacts.

A case study, presented in the 8th Webinar of the StopTB Partnerships’ Focus Group on AI-based Imaging for TB, showed that Uganda uses about 700,000 Xpert tests annually. Digital chest X-rays with CAD4TB can significantly reduce the number of Xpert tests needed, with minimal reduction of diagnosed TB cases.