Google Research deploys AI models for tuberculosis screening
In brief
- Google Research deploys AI tools for tuberculosis screening in high-burden areas.
- Open models MedGemma and MedSigLIP can be adapted to local health system needs.
- Over 25,000 X-rays analyzed, cutting diagnostic delays to under 45 seconds.
Google Research unveils AI tools designed to improve tuberculosis (TB) screening in high-burden areas. According to the WHO, TB kills 1.23 million people annually, with millions of cases undetected due to lack of access to radiologists or suitable infrastructure.
In 2019, the team demonstrated that deep learning algorithms could interpret chest X-rays with accuracy matching experienced radiologists. An independent evaluation by the Stop TB Partnership in 2021 confirmed these results. Since then, local partnerships have led to the certification and deployment of these tools in 40 healthcare facilities across six countries. Over 25,000 X-rays have been analyzed, reducing diagnostic delays from weeks to under 45 seconds.
To scale impact, Google Research has open-sourced its models (MedGemma, MedSigLIP) and embeddings, allowing local health systems to adapt them to their needs. Acoustic solutions (HeAR) complement the offering for areas without imaging access. These initiatives aim to democratize screening while respecting data sovereignty.
What it changes for you
- HealthAI tools enable rapid tuberculosis screening in under-equipped areas, cutting diagnostic delays from weeks to under 45 seconds.
This summary gives you the essentials. The full article, its figures and quotes are on Google Research.
Read the original articleAI-generated summary of the source, checked by rules. How we write · Report an error