How AI is revolutionising Indonesia’s Healthcare

Indonesia, the world’s third-highest TB burden country, has made remarkable progress against TB. Yet, over the last decade, the infectious disease has remained among the top causes of death in the nation. In a country of over 250 million, more than 950,000 are estimated to have developed TB. In addition, more than 50000 people are considered to be missing people with TB. Of these, a little over 47,000 are children.

In 2018, an MoU was signed between Indonesia and the Netherlands’ Ministry of Health departments. It included a collaboration between them to combat the rise of TB. This became the ground on which Delft Imaging joined the country’s effort against TB.

In 2020, we signed a Letter of Intent with Indonesia’s Ministry of Health on TB control using CAD4TB software. The government formally registered the software to support the efforts. Later that year, in collaboration with Fullerton Health, e-Tirta medical centres and the Ministry of Health, the CAD4TB was installed in several prisons and a factory in Jakarta. If any symptoms were found among the prisoners screened, the chest X-ray was taken to a mobile van equipped with CAD4TB. The CAD4TB box allows offline use with the data stored locally instead of being dependent on internet availability.

An optimised CAD4TB threshold was set with our partners, which defined whether a CXR was ‘abnormal’ and presumptive of TB. In such cases, an Xpert MIB/RIP test was conducted to confirm the TB potential. About a thousand prisoners were screened using CAD4TB. The pilot project reflected the potential to improve TB case finding in Indonesia. It concluded that not only was CAD4TB in line with the radiologist review, but it also reduced the time and cost of TB diagnosis.

In early 2022, in collaboration with e-Tirta medical centres and the Ministry of Health, the TB efforts were scaled up. In the first quarter of the year, over 50 thousand prisoners were screened. Later that year, the efforts were scaled up further with the partners. CAD4TB in 16 mobile clinics covered 25 public health centres and supported screening 110,000 people.

Alongside TB screening activities, we also collaborated with the Faculty of Medicine of Universitas Indonesia to research the use of AI within the nation. Three MoUs were signed. The first was on AI for TB screening (using CAD4TB), the second was on AI for COVID-19 screening, and the third MoU was on AI for retinal diseases (using RetCAD).

Making a difference

As a long-time partner to Indonesia’s TB screening efforts, researchers studied the use of CAD4TB among Indonesians living with Diabetes Mellitus (PLWD). This was particularly important as Diabetes Mellitus is a significant risk factor for TB. The study showed the potential use of CAD4TB as a triage tool for TB screening in PLWD, significantly reducing the need for microbiological examination.