arXiv 2509.098692025-09-11cs.CVcs.AI

用于鲁棒且可泛化的可变形图像配准的代理监督

Surrogate Supervision for Robust and Generalizable Deformable Image Registration

Department of Electrical and Computer Engineering, Vanderbilt University（维斯尼尔大学电气与计算机工程系）；Department of Radiology and Radiological Science, Johns Hopkins Medical School（约翰霍普金斯医学学校放射学与放射科学系）；Image Analysis and Communications Laboratory in the Department of Electrical and Computer Engineering, Johns Hopkins University（约翰霍普金斯大学电气与计算机工程系图像分析与通信实验室）；Center for Cognitive Medicine, Department of Psychiatry and Behavioral Science, Vanderbilt University Medical Center（维斯尼尔大学医学中心认知医学中心）；University of Pittsburgh, School of Medicine（匹兹堡大学医学院）

Yihao Liu, Junyu Chen, Lianrui Zuo, Shuwen Wei, Brian D. Boyd, Carmen Andreescu, Olusola Ajilore, Warren D. Taylor, Aaron Carass, Bennett A. Landman

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出代理监督训练范式，通过解耦输入域与监督域提升可变形图像配准网络对伪影、视野不匹配和模态差异的鲁棒性与泛化能力。

摘要：

目的：基于深度学习的可变形图像配准已达到较高精度，但仍对输入图像特征的变化（如伪影、视野不匹配或模态差异）较为敏感。我们旨在开发一种通用的训练范式，以提升配准网络的鲁棒性和泛化能力。方法：我们引入了代理监督，通过将估计的空间变换应用于代理图像，将输入域与监督域解耦。这使得在异质输入上训练成为可能，同时确保在相似性定义明确的域中计算监督。我们通过三个代表性应用评估了该框架：伪影鲁棒的大脑MR配准、与掩码无关的肺部CT配准以及多模态MR配准。结果：在各项任务中，代理监督对输入变化（包括不均匀场、不一致的视野和模态差异）表现出强大的抗性，同时在精心整理的数据上保持高性能。结论：代理监督提供了一个原则性框架，用于训练鲁棒且可泛化的基于深度学习的配准模型，且不增加复杂性。意义：代理监督为更鲁棒和可泛化的医学图像配准提供了一条实用途径，使其能够在多样化的生物医学成像场景中广泛应用。

英文摘要：

Objective: Deep learning-based deformable image registration has achieved strong accuracy, but remains sensitive to variations in input image characteristics such as artifacts, field-of-view mismatch, or modality difference. We aim to develop a general training paradigm that improves the robustness and generalizability of registration networks. Methods: We introduce surrogate supervision, which decouples the input domain from the supervision domain by applying estimated spatial transformations to surrogate images. This allows training on heterogeneous inputs while ensuring supervision is computed in domains where similarity is well defined. We evaluate the framework through three representative applications: artifact-robust brain MR registration, mask-agnostic lung CT registration, and multi-modal MR registration. Results: Across tasks, surrogate supervision demonstrated strong resilience to input variations including inhomogeneity field, inconsistent field-of-view, and modality differences, while maintaining high performance on well-curated data. Conclusions: Surrogate supervision provides a principled framework for training robust and generalizable deep learning-based registration models without increasing complexity. Significance: Surrogate supervision offers a practical pathway to more robust and generalizable medical image registration, enabling broader applicability in diverse biomedical imaging scenarios.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。