用隐式神经表示让图像配准自动调参，省去反复训练
cIDIR: Conditioned Implicit Neural Representation for Regularized Deformable Image Registration
- 基于隐式神经网络建模形变场，可微分支持灵活正则化
- 在DIR-LAB数据集上实现高精度与强鲁棒性配准
- 一次训练适配多种正则化参数，适合医学图像分析场景
正则化在可变形图像配准（DIR）中至关重要，能确保估计的形变矢量场（DVF）平滑、物理合理且解剖一致。然而，基于学习的DIR框架中精细调整正则化参数计算成本高，常需多次训练迭代。为此，我们提出cIDIR，一种基于隐式神经表示（INRs）的新框架，将注册过程条件化于正则化超参数。不同于传统方法需为每组超参数重新训练，cIDIR在超参数先验分布上训练后，仅通过分割掩码作为观测值即可优化正则化参数。此外，cIDIR建模连续可微的DVF，可通过自动微分无缝集成先进正则化技术。在DIR-LAB数据集上的评估显示，cIDIR在全数据集上均表现出高精度与强鲁棒性。
原文摘要 · Abstract (English)
Regularization is essential in deformable image registration (DIR) to ensure that the estimated Deformation Vector Field (DVF) remains smooth, physically plausible, and anatomically consistent. However, fine-tuning regularization parameters in learning-based DIR frameworks is computationally expensive, often requiring multiple training iterations. To address this, we propose cIDI, a novel DIR framework based on Implicit Neural Representations (INRs) that conditions the registration process on regularization hyperparameters. Unlike conventional methods that require retraining for each regularization hyperparameter setting, cIDIR is trained over a prior distribution of these hyperparameters, then optimized over the regularization hyperparameters by using the segmentations masks as an observation. Additionally, cIDIR models a continuous and differentiable DVF, enabling seamless integration of advanced regularization techniques via automatic differentiation. Evaluated on the DIR-LAB dataset, $\operatorname{cIDIR}$ achieves high accuracy and robustness across the dataset.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。