发表机构
Raidium（雷迪厄姆）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出Curia-MAE，通过改进MAE预训练目标，在30万张CT和MRI图像上预训练的多模态多解剖区域模型，可提升冻结编码器下的3D医学图像分割性能，尤其适用于标注数据稀缺的病灶任务。
AI 中文摘要
放射学基础模型学习可迁移表征，通过仅在冻结编码器顶部训练少量层即可适配新任务。然而，这类模型的评估中缺乏3D分割等密集预测任务的相关内容，且在编码器保持冻结的情况下，预训练模型的性能仍不及从头训练的基准模型nnU-Net。为缩小这一差距，本文在卷积掩码自编码器（MAE）预训练基础上，引入了鲁棒重建目标、特征正则化器以及局部-全局相似性目标。基于该方法，本文提出了Curia-MAE，这是一种在涵盖大量解剖部位的30万张CT和MRI图像上预训练的多模态、多解剖区域MAE模型。在8个聚焦解剖结构和病灶的分割基准测试中，Curia-MAE在编码器冻结的情况下，较强大的MAE基线模型提升了性能；在全微调设置下仍具备竞争力，且在标注数据稀缺的病灶分割任务中表现更优。这些结果表明，单个冻结编码器可在多种分割任务中复用，降低了这类模型在临床工作流中的适配与部署成本。本文将公开其预训练模型权重。
英文摘要
Radiology foundation models learn transferable representations that can be adapted to new tasks by training only small layers on top of a frozen encoder. Dense prediction tasks such as 3D segmentation are, however, underrepresented in their evaluation, and, with the encoder kept frozen, pre-trained models still fall short of nnU-Net, the state-of-the-art reference trained from scratch. To close this gap we extend convolutional MAE pre-training with a robust reconstruction objective, a feature regularizer, and a local-global similarity objective. Using this method, we propose Curia-MAE, a multi-modal, multi-anatomy MAE model pre-trained on 300,000 CT and MRI images covering a large number of anatomical sites. On eight anatomy- and lesion-focused segmentation benchmarks, Curia-MAE improves frozen-encoder performance over a strong MAE baseline, while remaining competitive under full finetuning and superior on lesion tasks, where labeled data is scarce. These results indicate that a single frozen encoder can be reused across diverse segmentation tasks, reducing the cost of adapting and deploying such models in clinical workflows. Curia-MAE pre-trained model weights are made publicly available at https://huggingface.co/raidium/Curia-MAE.
CommentsAccepted at ECCV 2026 Workshop AI4M3D
