Kikohozi Classifier​

The Kikohozi Classifier is an AI-powered cough audio analysis tool designed specifically for Tanzania to support the diagnosis and monitoring of respiratory diseases such as tuberculosis (TB), asthma, and chronic obstructive pulmonary disease (COPD). With a focus on analyzing cough sounds, the tool can help healthcare providers identify these conditions more quickly, non-invasively, and at lower cost.

Respiratory diseases remain a major cause of illness and death in Tanzania, where diagnostic resources are limited and often difficult to access. The Kikohozi Classifier aims to bridge this gap by offering a simple, affordable, and scalable solution that can strengthen early detection, improve treatment monitoring, and ultimately enhance health outcomes.

The tool is being developed for use in primary care facilities and community settings, particularly in semi-urban and rural areas where access to advanced diagnostic equipment is limited. The tool leveraging smartphones, which are increasingly available even in low-resource areas to offer a practical solution adapted to local realities.

The long-term goal is to make the Kikohozi Classifier accessible through a smartphone and web-based application, empowering health workers—and eventually the public—to assess cough-related illnesses quickly and effectively. Beyond Tanzania, this innovation aligns with the Sustainable Development Goals (SDGs), especially SDG 3 (Good Health and Well-being), by advancing early detection of infectious and non-communicable respiratory diseases, reducing premature deaths, and contributing to universal health coverage.