发表机构
Mannheim Institute for Intelligent Systems in Medicine (MIISM), Medical Faculty Mannheim, Heidelberg University; Department of Radiation Oncology, Brigham and Women’s Hospital, Dana-Farber Cancer Institute, Harvard Medical School; Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center Mannheim, Medical Faculty Mannheim, Heidelberg University; Department of Otolaryngology, Head and Neck Surgery, Campus Klinikum Bielefeld Mitte, University Hospital OWL of Bielefeld University; Interdisciplinary Center for Scientific Computing (IWR), Heidelberg University; Central Institute for Computer Engineering (ZITI), Heidelberg University; CZS Heidelberg Center for Model-Based AI, Heidelberg University（曼海姆医学智能系统研究所（MIISM），海德堡大学曼海姆医学院; 布莱根妇女医院放射肿瘤学系，达纳-法伯癌症研究所，哈佛医学院; 海德堡大学曼海姆医学院曼海姆大学医学中心耳鼻咽喉头颈外科; 比勒费尔德大学OWL大学医院比勒费尔德市中心校区耳鼻咽喉头颈外科; 海德堡大学跨学科科学计算中心（IWR）; 海德堡大学中央计算机工程研究所（ZITI）; 海德堡大学CZS基于模型的人工智能中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对软组织点云配准中对应估计难题，提出DINE框架，通过学习统计先验增强基于距离的配准，应用于两个主干并采用两阶段策略，实验表明其能降低平均倒角距离，提高对变形和噪声的鲁棒性，凸显全局变形合理性的重要性。
AI 中文摘要
非刚性点云配准是软组织形状分析的核心，但大变形、噪声和离群值使对应估计具有挑战性。大多数基于学习的方法依赖局部目标，如倒角距离，虽鼓励逐点接近，但不约束预测变形场的全局合理性。我们用DINE解决此限制，它是一个最大后验框架，通过对位移向量场的学习统计先验增强基于距离的配准。DINE应用于两个配准主干，采用两阶段策略：第一阶段模型用倒角距离训练，其预测变形场用于估计先验，然后用组合距离和负对数先验目标细化模型。我们比较全场PCA高斯先验和逐向量归一化流先验。在DeformedTissue和SynBench上的实验表明，在变形和损坏情况下平均倒角距离更低。在DeformedTissue上，相对于相应的第一阶段主干，DINE - PCA在不同变形水平下将倒角距离降低约27 - 69%，对离群值的鲁棒性提高高达66%，对高斯噪声的鲁棒性提高83%。在SynBench上，在最小变形水平下改进不大，从中等变形到严重变形时提高约59 - 79%。这些结果表明全局变形合理性是可靠软组织点云配准的重要约束。
英文摘要
Non-rigid point cloud registration is central to soft-tissue shape analysis, but large deformations, noise, and outliers make correspondence estimation challenging. Most learning-based methods rely on local objectives such as Chamfer distance, which encourage point-wise proximity but do not constrain the global plausibility of the predicted deformation field. We address this limitation with DINE, a maximum a posteriori framework that augments distance-based registration with a learned statistical prior over displacement vector fields. DINE is applied to two registration backbones, Robust-DefReg and DefTransNet, using a two-stage strategy: a first-stage model is trained with Chamfer distance, its predicted deformation fields are used to estimate a prior, and the model is then refined with a combined distance and negative log-prior objective. We compare a full-field PCA Gaussian prior with a per-vector normalizing-flow prior. Experiments on DeformedTissue and SynBench show lower mean Chamfer distance under deformation and corruption. On DeformedTissue, DINE-PCA reduces Chamfer distance by approximately 27--69\% relative to the corresponding Stage-1 backbone across deformation levels, and improves robustness by up to 66\% for outliers and 83\% for Gaussian noise. On SynBench, improvements are modest at the smallest deformation levels and reach approximately 59--79\% from moderate to severe deformation. These results suggest that global deformation plausibility is an important constraint for reliable soft-tissue point cloud registration. (The code will be published soon.)
