Image: Jacob Creswell / StopTB


Tuberculosis is one of the leading health issues in Pakistan. More than 350.000 TB cases were notified in 2016. Pakistan is a highly populated country and since tuberculosis is a very infectious disease, it can spread quickly.

In addition, Pakistan is among the countries with the highest occurrence of multidrug-resistant tuberculosis. According to the WHO, delay in diagnoses is one of the major problems with this form of TB. Success rates of tuberculosis treatments are no less than 93%, so an effective and clear method to detect TB was required. This way, diagnoses, and treatment can follow up instantly.

The Delft Approach

No less than 41 CAD4TB boxes including 300,000 images were installed in hospitals and mobile clinics. Our Computer Aided Detection for Tuberculosis (CAD4TB) software quickly and accurately detects the likelihood of active TB in an X-ray image of a patient using revolutionary machine learning technology. This way, the TB case detection accelerates, and a patient can be diagnosed and start treatment within a day. The results are outstanding: every month 60,000 people are screened for tuberculosis.

In May 2017, training on CAD4TB in general, practical use of the software and troubleshooting and admin tools for the medical staff in Pakistan were carried out.

Making an Impact

Soon after Pakistan started using CAD4TB 35 more CAD4TB boxes were requested. They are planning to do more than 2 million TB screenings for the next three years. Delft is proud to say our CAD4TB software can make an impact in the fight against tuberculosis in Pakistan.