stub Intel, Penn Medicine Conduct Largest Medical Federated Learning Study - Unite.AI
Connect with us

Healthcare

Intel, Penn Medicine Conduct Largest Medical Federated Learning Study

Published

 on

Image: Intel Labs

Intel Labs and the Perelman School of Medicine at the University of Pennsylvania (Penn Medicine) have announced the results of the largest medical federated learning study. The joint research study used machine learning (ML) and artificial intelligence (AI) to help international healthcare and research institutions identify malignant brain tumors. 

The research was published in Nature Communications

An Unprecedented Study

The study involved an unprecedented dataset examined from 71 institutions spread across six continents, and its results demonstrated the ability to improve brain tumor detection by 33%. 

Jason Martin is principal engineer at Intel Labs. 

“Federated learning has tremendous potential across numerous domains, particularly within healthcare, as shown by our research with Penn Medicine,” Martin said. “Its ability to protect sensitive information and data opens the door for future studies and collaboration, especially in cases where datasets would otherwise be inaccessible. Our work with Penn Medicine has the potential to positively impact patients across the globe and we look forward to continuing to explore the promise of federated learning.”

Data Accessibility in Healthcare

Data accessibility is a major challenge in healthcare, with state and national data privacy laws making it hard to conduct medical research and data at scale without compromising patient health infromation. Thanks to confidential computing, the federated learning hardware and software from Intel comply with data privacy concerns and preserve data integrity.

The teams processed high volumes of data in a decentralized system using Intel federated learning technology along with Intel Software Guard Extensions (SGX), which help remove data-sharing barriers. The system also addresses privacy concerns by maintaining raw data inside the data holders’ compute infrastructure. Model updates computed from the data can only be sent to a central server or aggregator. The data itself cannot be sent. 

Rob Enderle is principal analyst at Enderle Group. 

“All of the computing power in the world can’t do much without enough data to analyze,” said Enderle. “This inability to analyze data that has already been captured has significantly delayed the massive medical breakthroughs AI has promised. This federated learning study showcases a viable path for AI to advance and achieve its potential as the most powerful tool to fight our most difficult ailments.”

Spyridon Bakas, PhD, is an assistant professor of Pathology & Laboratory Medicine, and Radiology, at the Perelman School of Medicine at the University of Pennsylvania. 

“In this study, federated learning shows its potential as a paradigm shift in securing multi-institutional collaborations by enabling access to the largest and most diverse dataset of glioblastoma patients ever considered in the literature, while all data are retained within each institution at all times,” said Bakas. “The more data we can feed into machine learning models, the more accurate they become, which in turn can improve our ability to understand and treat even rare diseases, such as glioblastoma.”

It’s critcial for researchers to have access to large amounts of medical data to advance treatments. But this amount of data is usually too much for one facility. With the new study, researchers are closer to unlocking multisite data silos to advance federated learning at scale. These advancements could bring on many benefits like the early detection of disease. 

Alex McFarland is a Brazil-based writer who covers the latest developments in artificial intelligence & blockchain. He has worked with top AI companies and publications across the globe.