Medical AI Needs Federated Learning, So Will Every Industry

By
The Rhino Team
No items found.
September 15, 2021

A study published in Nature Medicine highlights the effectiveness of federated learning in developing AI models that generalize across multiple healthcare institutions without compromising data privacy. The research involved 20 hospitals across five continents collaborating to train a neural network, named EXAM, designed to predict the level of supplemental oxygen a COVID-19 patient might require within 24 to 72 hours of arrival. Federated learning allowed each institution to train the model locally on their data, sharing only model updates rather than raw data. This approach preserved patient confidentiality while producing a robust, generalizable AI model. The success of this study suggests that federated learning can be applied beyond healthcare, benefiting industries like energy, finance, and manufacturing, where data privacy and security are paramount.

For more details, read the full article on NVIDIA’s blog.

Build your Federated Network