AI Boosts Healthcare and Science

Last week at the Lake Nona Impact Forum, I talked about how AI could revolutionize healthcare and science. With our latest AI advancements, we're on the cusp of making healthcare more accessible, personalized, and effective for everyone, while also speeding up scientific breakthroughs. Here's a rundown of what we've been up to, our collaborations to bring AI into healthcare worldwide, and our newly announced AI co-scientist.

AI is making accurate health information more accessible

When folks need health info, they often turn to Google first. That's why we're committed to making sure everyone can get high-quality, relevant health info when they need it most. With Google Lens, you can snap a pic to search for skin conditions that look like what you're seeing on your own skin. Over on YouTube, we've been testing AI tools with health creators and groups like the Cleveland Clinic to help them put out top-notch, authoritative content more easily.

For healthcare providers, we've rolled out MedLM and Search for Healthcare, which can answer medical queries. These tools are available on the Google Cloud Vertex AI platform, helping doctors make better decisions and ensuring patients get the precise care they need. We're also working hard to make sure the health-related content from our language models is as accurate and fact-based as possible.

Generative AI is paving the way for personalized healthcare

Thanks to progress in multimodality and conversational AI, we're rethinking how we can tailor patient care to each individual, focusing especially on preventive healthcare.

Healthcare is all about different types of data, from X-rays to digital records. Building on our MedLM research, we've created Med-Gemini, a new set of models for healthcare that leverage Gemini's strong multimodal and reasoning skills, fine-tuned with de-identified medical data. In our published research, Med-Gemini scored 91.1% accuracy on U.S. medical exam-style questions, and we've shown how these models can effectively interpret 3D scans or tackle complex clinical questions.

We're also exploring how AI can act as a conversational diagnostic tool in clinics with our Articulate Medical Intelligence Explorer (AMIE). This AI system is designed to take a "clinical history" and ask smart questions to help reach a differential diagnosis, all while handling discussions with empathy, even in specialized fields.

Mobile and wearable tech is another exciting area where generative AI can offer personalized health and wellness insights, using data like step count and heart rate. We've developed the Personal Health Large Language Model, a fine-tuned version of Gemini, which can analyze sensor data to provide insights and recommendations on an individual's sleep and fitness patterns.

AI is improving health outcomes globally

Catching diseases early can make a huge difference in health outcomes. Over the last ten years, we've harnessed AI's imaging and diagnostic powers to develop models that help detect diseases like breast cancer, lung cancer, and diabetic retinopathy. Through partnerships, we're now scaling these solutions in clinical settings, helping more patients get timely and accurate screenings. This is especially impactful in areas with limited medical resources and fewer specialists. In the coming decade, our partners in India and Thailand plan to provide 6 million free diabetic retinopathy screenings, while Apollo Radiology International will use our AI models to offer 3 million free screenings for tuberculosis, lung cancer, and breast cancer in India. We're also working on maternal health in Africa, developing an ML model for cardiotocography to predict fetal well-being, and looking at how it can be used in resource-limited settings.

We're also building the tech foundations to make healthcare more accessible. Our Health AI Developer Foundations include open-weight models and resources to help developers build AI models for healthcare more efficiently. Solutions powered by our Open Health Stack (OHS) — a set of open-source tools that make it easier for developers to create next-gen digital health solutions — are already being used in Africa, South Asia, and Southeast Asia to support millions of patients.

AI is accelerating scientific discovery

Healthcare is built on science, and AI's knack for synthesizing information and performing complex reasoning tasks is helping us explore new ways to boost scientific and biomedical discovery. Our AI co-scientist, a multi-agent system based on Gemini 2.0, is designed to be a collaborative tool for scientists. It aims to uncover new knowledge, help formulate novel research hypotheses and proposals based on existing evidence, and tailor them to specific research goals. It's already shown promise in areas like drug repurposing for acute myeloid leukemia, suggesting new treatment targets for liver fibrosis, and explaining mechanisms of horizontal gene transfer related to antimicrobial resistance — each a complex challenge in its own right.

We're seeing the huge potential of AI to push forward science and make healthcare more personalized and accessible. The cycle of turning research breakthroughs into real-world impact is speeding up and growing wider. We're committed to doing this responsibly, working with global partners, and sharing our research — in 2024 alone, we published over 50 papers on cutting-edge health research, and we've just released our 2025 Health Impact Report. We believe AI will keep advancing healthcare and science, benefiting billions of people.