Personalized Federated Learning (pFL) effectively addresses the challenge of statistical heterogeneity in traditional Federated Learning (FL), with feature alignment methods (e.g., FedProto) standing out due to their communication efficiency and model-agnostic design, making them practically viable in real-world non-IID scenarios. These methods directly align class-level features across clients without requiring model parameter transmission. However, they represent each class as a holistic prototype, which limits the diversity and expressiveness of shared features. This restriction hampers the model's ability to generalize across clients and impedes personalized adaptation, as clients lack sufficient semantic components to reconstruct discriminative features tailored to their local data distributions. To overcome these limitations, we propose Federated Visual Primitive Learning (FedVPL), a novel framework comprising two key components: (1) Visual Primitive Space Sharing, which decomposes class-level features into semantically meaningful and reusable visual primitives, enabling cross-client and cross-class sharing to enrich feature diversity and decouple communication cost from the number of classes, significantly improving efficiency; and (2) Text-Guided Semantic Alignment, a parameter-free personalization mechanism that leverages external language priors to align shared primitives with client-specific semantics, without requiring additional communication overhead. Extensive experiments across diverse non-IID benchmarks demonstrate that FedVPL substantially outperforms state-of-the-art baselines, achieving up to a 5.62% improvement in accuracy, reducing communication overhead by at least 12.5x, and effectively addressing generalization and personalization challenges in heterogeneous federated environments.

In recent years, score-based diffusion models have emerged as effective tools for estimating score functions from empirical data distributions, particularly in integrating implicit priors with inverse problems like CT reconstruction. However, score-based diffusion models are rarely explored in challenging tasks such as metal artifact reduction (MAR). In this paper, we introduce a Bi-Constraints Diffusion Model for Metal Artifact Reduction (BCDMAR), an innovative approach that enhances iterative reconstruction with a conditional diffusion model for MAR. This method employs a metal artifact degradation operator in place of the traditional metal-excluded projection operator in the data-fidelity term, thereby preserving structure details around metal regions. However, score-based diffusion models tend to be susceptible to grayscale shifts and unreliable structures, making it challenging to reach an optimal solution. To address this, we utilize a pre-corrected image as a prior constraint, guiding the generation of the score-based diffusion model. By iteratively applying the score-based diffusion model and the data-fidelity step in each sampling iteration, BCDMAR effectively maintains reliable tissue representation around metal regions and produces highly consistent structures in non-metal regions. Through extensive experiments focused on metal artifact reduction tasks, BCDMAR demonstrates superior performance over other state-of-the-art unsupervised and supervised methods, both quantitatively and qualitatively.

Existing deep learning (DL)-based magnetic resonance imaging (MRI) retrospective motion correction (MoCo) models are typically task-specific, which makes them challenging to generalize to different scenarios w.r.t motions, modalities, planes, and scanner centers. This limitation occurs since the motions of each patient vary, and collecting diverse paired/unpaired motion data is generally costly and infeasible. To deal with this problem, we propose the Equivariant Imaging Prior (EIP) framework to generalize the MoCo tasks toward various scenarios.In this paper, the traditional MRI MoCo tasks, specifically for the multi-scenarios, can be treated as a mask-varying compressed sensing self-supervised problem for MRI reconstruction with corrupted k-space data.To the best of our knowledge, this framework is the first attempt to handle multiple MRI MoCo scenarios with one single DL model. Specifically, stochastic subsampling and modality augmentation are employed for data preparation. Then, a domain generalization-friendly net is carefully designed and an equivariant imaging task is leveraged to learn the mapping from corrupted data to clean images. The experimental results show that the proposed EIP framework achieves impressive adaptability across generalizable MoCo tasks, including but not limited to multi-motion, multi-modality, multi-center, and multi-plane. Furthermore, our EIP demonstrates similar or superior performance to several state-of-the-art models trained in a supervised manner, extending to even motion estimation on the multi-coil raw data. The code is available:https://github.com/wangzhiwen-scu/EIP4MoCo.

Metal artifacts degrade CT image’s quality. Recently, some deep learning-based metal artifact reduction (MAR) methods have been developed. Supervised MAR methods don’t perform well in clinical due to the domain gap between simulated and clinical data. Although this problem can be avoided in an unsupervised way, severe artifacts cannot be well suppressed. Semi-supervised MAR methods can alleviate the domain gap problem. However, the existing ones are usually accompanied by boosted model scale, which is challenging for optimization. In this paper, we propose a novel semi-supervised framework for MAR, termed SemiMAR. First, we only use one generator to learn the clean part, instead of multiple encoders and decoders disentangling artifacts. Thus, the model naturally becomes much smaller. To recover more tissue details, the advanced dual-domain MAR network knowledge is distilled into our model in both the image domain and latent feature space. Extensive experiments demonstrate the efficiency and robustness of our model.

Metal artifacts lead to CT imaging quality degradation. With the success of deep learning (DL) in medical imaging, a number of DL-based supervised methods have been developed for metal artifact reduction (MAR). Nonetheless, fully-supervised MAR methods based on simulated data do not perform well on clinical data due to the domain gap. Although this problem can be avoided in an unsupervised way to a certain degree, severe artifacts cannot be well suppressed in clinical practice. Recently, semi-supervised metal artifact reduction (MAR) methods have gained wide attention due to their ability in narrowing the domain gap and improving MAR performance in clinical data. However, these methods typically require large model sizes, posing challenges for optimization. To address this issue, we propose a novel semi-supervised MAR framework. In our framework, only the artifact-free parts are learned, and the artifacts are inferred by subtracting these clean parts from the metal-corrupted CT images. Our approach leverages a single generator to execute all complex transformations, thereby reducing the model's scale and preventing overlap between clean part and artifacts. To recover more tissue details, we distill the knowledge from the advanced dual-domain MAR network into our model in both image domain and latent feature space. The latent space constraint is achieved via contrastive learning. We also evaluate the impact of different generator architectures by investigating several mainstream deep learning-based MAR backbones. Our experiments demonstrate that the proposed method competes favorably with several state-of-the-art semi-supervised MAR techniques in both qualitative and quantitative aspects.