TEDS-TB: ICT tools to support the early diagnosis of tuberculosis in Peru

The project responded to the need to expand the training of non-microbiological healthcare staff in the diagnosis of prevalent and neglected diseases, using tuberculosis as a test case to validate the system in a real-world setting. The usual diagnostic method, recommended by the WHO, required manually identifying and counting the bacilli in sputum, a slow procedure prone to human error.

The objectives of the project were:

• To develop open-source early-warning software for tuberculosis based on artificial intelligence and computer vision.
• To train the software in a controlled environment and validate it in a real environment.
• To integrate the software into a web portal that would allow remote online analysis.

The project was carried out in five phases: software development, training with a bank of 1,600 images, validation with 1,000 tuberculosis cases in San Juan de Lurigancho, integration into a bilingual web portal with a user manual, and transfer of the project to the Instituto de Medicina Tropical Alexander von Humboldt.

CTIC carried out the project with funding from the Spanish Agency for International Development Cooperation (AECID).