UniH$^3$：统一分层同质性与异质性用于全能医学图像恢复
UniH$^3$: Unifying Hierarchical Homogeneity and Heterogeneity for All-in-One Medical Image Restoration
浏览论文内容
中文总结 AI 辅助
提出UniH3框架，通过分层同质性记忆和异质性平衡器统一医学图像中的同质性与异质性，实现全能医学图像恢复，在两个大规模基准上达到最先进性能。
中文摘要 AI 辅助
全能医学图像恢复（MedIR）旨在使用单一通用模型处理跨模态和退化类型的多样任务。现有方法通常优先考虑建模任务间异质性（例如，不同的数据分布和退化类型）。然而，它们很大程度上忽视了医学图像中固有的同质性，例如模态内部和跨模态广泛共享的解剖结构，这些结构可用于简化模型训练并提高泛化能力。为此，我们提出了UniH3，一种新颖的框架，统一了分层同质性与异质性，用于全能医学图像恢复。具体而言，为全面利用同质性，我们引入了分层同质性记忆（H2M）模块，该模块在训练期间从高质量图像中逐步提炼任务内和任务间的同质性先验，并根据输入自适应检索最相关的先验以指导恢复。这些检索到的先验随后通过高效的基于同质性的注意力（HGA）机制注入到恢复流程中。此外，为全面处理异质性，我们设计了分层异质性平衡器（H2B），以在优化过程中缓解任务间和任务内的冲突，促进平衡且有效的多任务学习。在两个大规模基准测试MedIR-2D-500K和MedIR-3D-3K上的广泛实验表明，UniH3在全能和单任务医学图像恢复方面均达到了最先进的性能。我们希望这项工作能够建立一个强基准，并推动通用医学图像恢复模型的发展。代码可在以下网址获取：https://this https URL。
英文摘要
All-in-One medical image restoration (MedIR) aims to address diverse tasks across modalities and degradation types using a single universal model. Existing methods typically prioritize modeling inter-task heterogeneity (e.g., distinct data distributions and degradation types). However, they largely neglect the inherent homogeneity present in medical images, such as widely shared anatomical structures within and across modalities, which can be leveraged to ease model training and improve generalization. To this end, we propose UniH3, a novel framework that Unifies Hierarchical Homogeneity and Heterogeneity for all-in-one medical image restoration. Specifically, to comprehensively exploit homogeneity, we introduce a Hierarchical Homogeneity Memory (H2M) module that progressively distills intra- and inter-task homogeneity priors from high-quality images during training, and adaptively retrieves the most relevant priors tailored to the input for guided restoration. These retrieved priors are then injected into the restoration pipeline via an efficient Homogeneity-Guided Attention (HGA) mechanism. Furthermore, to comprehensively address heterogeneity, we design a Hierarchical Heterogeneity Balancer (H2B) that mitigates both inter- and intra-task conflicts during optimization, facilitating balanced and effective multi-task learning. Extensive experiments on two large-scale benchmarks, MedIR-2D-500K and MedIR-3D-3K, demonstrate that UniH3 achieves state-of-the-art performance on both all-in-one and single-task medical image restoration. We hope this work establishes a strong benchmark and advances the development of general-purpose medical image restoration models. Code is available at https://github.com/Yaziwel/UniH3.
发表机构
- School of Biological Science and Medical Engineering, Beihang University（北京航空航天大学生物科学与医学工程学院）
- Department of Biomedical Engineering, Tsinghua University（清华大学生物医学工程系）
机构由 AI 辅助整理，请以论文原文为准。
