Fortifying Medical Diagnosis: Federated Learning Augmented by Blockchain for Enhanced Privacy, Security, and Robustness

   page       attach   
Lorenzo Cassano  •  Jacopo D’Abramo
sommario

In response to the critical need for safeguarding privacy and security within the medical sector,
this project capitalizes on recent advancements in artificial intelligence (AI). Leveraging the
innovative architecture of Federated Learning pioneered by Google, combined with the advantages
offered by blockchain technology, our objective is to develop a framework that prioritizes
trust, reliability, and ethical conduct throughout the learning process within a Federated Learning
system.
Specifically, our focus lies in employing this framework to detect brain tumor cancer through
an Artificial Neural Network (ANN), thereby simulating a realistic scenario where each participant
in the Federated Learning process represents a distinct hospital with its own repository of sensitive
patient data. By integrating Federated Learning with blockchain technology, we aim to establish
a robust infrastructure capable of ensuring the confidentiality, integrity, and security of patient
information while fostering collaborative learning among distributed healthcare institutions.

parole chiave
Federated learning, Blockchain, ETH, CNN, Neuroimaging
prodotti