By 
Doctors spend years training to get to where they are today. Even then, the learning is never over. Doctors need to stay on top of new developments to ensure they can deliver the best healthcare for their patients. But doctors are humans too, which means sometimes they can make mistakes, which is why Google is hoping to leverage its Gemini AI for the medical industry.
Jeff Dean, Google’s Chief Scientist, Google DeepMind and Google Research, shared a post on X in which he revealed Med-Gemini. This is a variant of Google’s Gemini AI that has been tuned for medical purposes. Based on the examples shown, it sounds like it is more or less Gemini that some of you are familiar with. The difference is that it has been tuned to answer and analyze questions related to medicine.
Med-Gemini is built on Gemini-1.0/1.5 & can be easily adapted to new medical modalities with custom encoders. We showcase Med-Gemini’s promise in accurate multimodal dialogue🗣️🖼️ with examples of high-quality conversations about radiology & dermatology images, noting that… pic.twitter.com/VSdU3fvijk
— Alan Karthikesalingam (@alan_karthi) April 30, 2024
One of the examples shown involved a student asking the AI for additional details about a surgical video. It was also able to answer questions about the surgical procedure and what tools are used. It also provided advice on how to prevent injuries during the surgery. In the demo video above, Google’s Alan Karthikesalingam also showed an example of how an image of a medical condition can be shown to the AI, and have the AI analyze it and offer a potential diagnosis.
Keep in mind that Med-Gemini is currently not available for use yet. Google has published a research paper that merely highlights some of its capabilities and methodologies.
Comments