Projection Guided Personalized Federated Learning for Low Dose CT Denoising

2026-03-12

AI Summary

This work addresses the challenge of denoising low-dose computed tomography (LDCT) images corrupted by scanner-specific noise and artifacts, which existing federated learning methods struggle to disentangle from patient anatomy in image space. To overcome this limitation, the authors propose ProFed, a novel framework that achieves dual personalization in projection space for the first time. ProFed employs anatomy-aware and protocol-aware networks to model individual variations, enforces physical consistency through a multi-constrained projection loss, and introduces an uncertainty-guided selective client aggregation mechanism. Evaluated on the Mayo Clinic 2016 dataset, ProFed achieves peak signal-to-noise ratios (PSNR) of 42.56 dB and 44.83 dB using CNN and Transformer backbones, respectively, significantly outperforming eleven federated learning baselines and surpassing SCAN-PhysFed by 1.42 dB.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Adversarial Attacks & Robustness
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Low-dose CT (LDCT) reduces radiation exposure but introduces protocol-dependent noise and artifacts that vary across institutions. While federated learning enables collaborative training without centralizing patient data, existing methods personalize in image space, making it difficult to separate scanner noise from patient anatomy. We propose ProFed (Projection Guided Personalized Federated Learning), a framework that complements the image space approach by performing dual-level personalization in the projection space, where noise originates during CT measurements before reconstruction combines protocol and anatomy effects. ProFed introduces: (i) anatomy-aware and protocol-aware networks that personalize CT reconstruction to patient and scanner-specific features, (ii) multi-constraint projection losses that enforce consistency with CT measurements, and (iii) uncertainty-guided selective aggregation that weights clients by prediction confidence. Extensive experiments on the Mayo Clinic 2016 dataset demonstrate that ProFed achieves 42.56 dB PSNR with CNN backbones and 44.83 dB with Transformers, outperforming 11 federated learning baselines, including the physics-informed SCAN-PhysFed by +1.42 dB.

Problem

• Low-dose CT
• federated learning
• noise separation
• personalization
• projection space

Innovation

• Projection Space Personalization
• Federated Learning
• Low-Dose CT Denoising
• Uncertainty-Guided Aggregation
• Multi-Constraint Projection Loss