FedProIn framework enhances federated learning for medical imaging

Researchers have developed FedProIn, a novel framework designed to improve federated learning in medical imaging by addressing client drift. This approach utilizes learnable class prototypes to capture shared semantic structures across heterogeneous clients, mitigating issues caused by variations in scanners, acquisition protocols, and patient populations. FedProIn introduces feature divergence loss and prototype contrastive loss to combat client drift and employs a normalized influence aggregation strategy to adaptively weight client prototypes based on their contribution to the global representation. Experiments on the HAM10000 and Matek-19 datasets show FedProIn achieving high accuracies, outperforming existing baselines in both IID and non-IID conditions. AI

IMPACT This research could lead to more robust and accurate AI models in healthcare by improving federated learning techniques for medical imaging.

RANK_REASON The cluster contains an academic paper detailing a new method for federated learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• client drift
• federated learning
• FedProIn
• HAM10000
• Matek-19
• Vaanathi Sundaresan PhD

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →