Skip to content
EVENT / TOPICHistorical

Google Research releases open-source medical AI models MedGemma and HAI-DEF for TB screening

1 reports1 sources10/23/2026 updated

Current overview (not historical)

AI overview

Google Research has released two open-source medical AI models, MedGemma and HAI-DEF, aimed at solving tuberculosis (TB) screening issues through localized deployment. MedGemma has advanced from basic classification to a multimodal clinical reasoning model, capable of answering natural language questions by integrating chest X-rays with clinical history. Crucially, it can run locally on devices without internet access and has achieved 10 million global downloads. HAI-DEF specifically focuses on TB detection and has secured CE MDR Class IIb regulatory approval, currently deployed in 40 medical institutions across countries like South Africa and Sierra Leone, screening over 25,000 chest X-rays. Both models emphasize running on local infrastructure to ensure patient data remains within national borders, supporting high-burden TB countries like Indonesia in achieving a target of 50 million annual screenings.

Written with AI from the reporting · updated 10/23/2026

Latest developmentOct 24, 02:51
Google Research deploys open foundation models like MedGemma for global TB screening
Explore the timeline, relations and evidence
Expand the detailed fact route and sources

Story route

A station-by-station route from the first public evidence onward; every station carries its own evidence and uncertainty.

  1. 1Unclassified fact stationOct 24, 00:00recorded Oct 24, 03:16single sourcerevised

    Google Research发布MedGemma和HAI-DEF开源模型

    Read original source ↗ Google ResearchReplay this moment in the explorer
    Evidence & revisions · 1 source records
    Subject：
    google
    Action：
    发布
    Object：
    MedGemma和HAI-DEF开源模型
    Occurred (fact side)：
    Oct 24, 00:00
    Google Research ↗Report pageRevision 2 · currently revised
    Published (valid side)：
    Oct 24, 02:51
    First seen (known side)：
    Oct 24, 03:16
    Fact stationSingle source · one-sided claimRevised

    Copy may have been revised later; time replay is bounded. Layer labels group reading only, not causation between layers.

Uncertainty marks flag where to check the sources; they are not proven causation.

Report timeline

Explore every side through dated source reports.

Oct 24
  1. Google ResearchSelectedFirst-party sourceRelated fact: single sourceSource report revised
    Google Research deploys open foundation models like MedGemma for global TB screening

    Report first recorded Oct 24, 03:16

    Google Research and Google for Health are deploying open foundation models, such as MedGemma, to democratize tuberculosis (TB) screening globally, addressing the scarcity of radiologists in high-burden settings. Initial work demonstrated deep learning algorithms could detect active pulmonary TB on chest radiographs with accuracy matching expert radiologists (2019). The model was ranked among top global TB screening algorithms by the Stop TB Partnership (2021), and a prospective multi-site validation in NEJM AI (2024) confirmed high sensitivity across diverse populations. In partnership with Nexus AI, the tool has received regulatory clearance (CE MDR Class IIb) and has been deployed in 40 medical institutions across countries like South Africa and Sierra Leone, screening over 25,000 chest X-rays.

    Read original source ↗

Event discussion

Reader opinions are hosted on GitHub Discussions; posting requires a GitHub account. Comments are not verified facts.

Open GitHub Discussions