Projection Guided Personalized Federated Learning for Low Dose CT Denoising
This addresses protocol-dependent noise in low-dose CT across institutions, enabling better personalized denoising without centralizing patient data, but it is incremental as it builds on existing federated learning methods.
The paper tackled the problem of low-dose CT denoising in federated learning by proposing ProFed, which personalizes in projection space to separate scanner noise from patient anatomy, achieving 44.83 dB PSNR with Transformers and outperforming baselines by up to +1.42 dB.
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.