Federated learning for biomedical applications
Closes 01 May, 2024
Issue Pre-order formJournal: International Journal of Sensors, Wireless Communications and Control
Guest editor(s): Ranjith R
Introduction
Federated learning, also known as distributed AI/machine learning, is a method that enables cooperative learning from big datasets owned by several parties without compromising the privacy of each person's raw data. FL is especially helpful when the needed information is not open source or easily accessible due to tactical or legal considerations. By using a collaborative model training technique without revealing sensitive data, it also aims to solve impending privacy and data governance challenges. With benefits for privacy and network traffic, federated learning (FL) develops models in the peripheral and distributes them without always transmitting data. By comparing their Artificial Intelligence (AI) models with those of other institutions without revealing patient information, hospitals and clinics might, for instance, enhance their AI models in medical research. Even if anonymized, it is best to keep data at the location where it was created and/or kept. In edge computing networks, federated learning can help with issues including non-IID training data, restricted communication, uneven contributions, privacy, and security. As is generally known, there is no longer a problem with data availability. If we wanted to work together, could we share data? Consider the medical industry and the Covid19 pandemic's impact. It would be remarkable to develop a model that can identify the illness from an X-ray, but there are challenges. A hospital in China cannot easily communicate its data with a hospital in India due to regulatory restrictions on privacy and law. The second is more realistic: even under the supposition that all hospital management worldwide would have given their approval, a massive infrastructure would still be required that would need to be rented out and managed. Where was it erected? Who oversees it? It is actually not practicable. However, there is a real need for sharing. In this view, a smart gadget replaces a sensor that just snaps a photo and sends it, processing information instantly and alerting the user if there are any issues. Federated learning enables you to propagate the learning of the single experience to all the other sensors if the device is trained on edge, that is, directly in the field, without having to repeatedly visit the data center to train the model. With regard to federated learning for biomedical applications, this special issue aims to address cutting-edge research directions and contributions.
Keywords
Federated learning, Artificial Intelligence, Tissue Engineering, Optimization, Sustainability, Sensors, Wireless communication.