arXiv 2603.040242026-07-08cs.CVcs.AI

体积方向扩散：在模糊医学图像分割的解剖学共识中锚定不确定性量化

Volumetric Directional Diffusion: Anchoring Uncertainty Quantification in Anatomical Consensus for Ambiguous Medical Image Segmentation

Department of Computer Science and Engineering University at Buffalo, SUNY Buffalo, NY, USA（计算机科学与工程系，布法罗大学，纽约州，布法罗，美国）

Chao Wu, Mahesh Bhosale, Kangxian Xie, Pouya Karimian, David Doermann, Mingchen Gao

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对模糊医学图像分割中不同专家边界描绘有差异的问题，提出体积方向扩散（VDD）框架，以粗略共识预测为锚点，学习方向扩散过程生成边界变化，实验表明该方法能在保持精度和结构一致性时改善不确定性分布对齐。

摘要：

模糊的3D医学图像分割通常涉及不同专家描绘不同但临床上合理的边界。对这种观察者间变异性进行建模需要在多样性和解剖学保真度之间仔细平衡：确定性模型保留连贯的体积结构，但将专家分歧归结为单个掩码，而随机生成模型可以产生多样的样本，但在从无结构化噪声生成完整的3D掩码时可能会引入不相连的组件或切片到切片的不一致。我们提出了体积方向扩散（VDD），这是一种先验锚定的扩散框架，它将随机生成从全掩码合成转移到残余边界探索。VDD使用粗略的共识预测作为解剖学锚点，并学习方向扩散过程，以在保留稳定的体积拓扑的同时，在模糊区域周围生成合理的边界变化。在包括LIDC-IDRI、KiTS21和ISBI 2015在内的三个多评估者数据集上的实验表明，VDD在保持有竞争力的分割精度和3D结构一致性的同时，改善了不确定性分布对齐。这些结果表明，先验锚定的残余扩散可以在不牺牲解剖学保真度的情况下，对临床上合理的专家分歧进行建模。

英文摘要：

Ambiguous 3D medical image segmentation often involves boundaries where different expert delineations are non-identical yet clinically plausible. Modeling such inter-observer variability requires a careful balance between diversity and anatomical fidelity: deterministic models preserve coherent volumetric structures but collapse expert disagreement into a single mask, while stochastic generative models can produce diverse samples but may introduce disconnected components or slice-to-slice inconsistency when generating full 3D masks from unstructured noise. We propose Volumetric Directional Diffusion (VDD), a prior-anchored diffusion framework that shifts stochastic generation from full-mask synthesis to residual boundary exploration. VDD uses a coarse consensus prediction as an anatomical anchor and learns a directional diffusion process to generate plausible boundary variations around ambiguous regions while preserving stable volumetric topology. Experiments on three multi-rater datasets, including LIDC-IDRI, KiTS21, and ISBI 2015, show that VDD improves uncertainty distribution alignment while maintaining competitive segmentation accuracy and 3D structural consistency. These results suggest that prior-anchored residual diffusion can model clinically plausible expert disagreement without sacrificing anatomical fidelity.

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。