This invention describes a system that predicts how healthy someone is in the short term by analyzing data from their wearable device, like a fitness tracker, combined with information from their health records. It uses artificial intelligence, specifically a random forest model, to sort users into categories. If someone is classified as healthy, they get a baseline health score, but if not, a different AI model calculates a more detailed score. Specifically, the claims focus on using data from an activity tracker, including heart rate, oxygen, activity levels, calories burned, and even manually entered information.
Why it matters: Filed before advanced AI, particularly large language models, significantly improved processing unstructured health records. Generating the crucial "embedding data" from patient records is now much more efficient.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI