发表机构
Faculty of Computing, Harbin Institute of Technology; College of Computer and Control Engineering, Northeast Forestry University; Case Western Reserve University（哈尔滨工业大学计算学部; 东北林业大学计算机与控制工程学院; 凯斯西储大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对现有模糊医学图像分割方法无法形成渐进式语义建模的问题，提出KANResDiff模型，通过独立时间编码与残差薛定谔桥实现阶段感知模糊性建模，在公开数据集上取得SOTA性能。
AI 中文摘要
模糊医学图像分割旨在提供一系列多样且合理的分割假设，但现有方法以固定且预定义的方式引入随机性，无法形成渐进式语义建模过程。为应对这些挑战，我们提出KANResDiff，通过柯尔莫哥洛夫-阿诺德网络学习局部残差扩散，从而在各阶段分配不同角色以实现模糊性建模。具体而言，我们提出独立时间编码，提供基于样条的时间嵌入而非多层感知机（MLP）的线性嵌入，这增强了推理阶段间的独立性，并为不同阶段分配渐进式语义角色；我们提出残差薛定谔桥，通过构建局部薛定谔桥注入带可学习权重的确定性残差先验，而非遵循手动设置，借助局部最优扩散路径实现灵活的确定-随机交互与阶段感知的模糊性建模。在两个公开数据集上的大量实验结果表明，KANResDiff在GED和HM-IoU指标上实现了SOTA性能，最大提升分别为16.8%和7.7%，同时在MDM指标上保持了有竞争力的性能。源代码可在指定URL获取。
英文摘要
Ambiguous medical image segmentation aims to provide a series of diverse but plausible segmentation hypotheses. However, existing methods introduce stochasticity in a fixed and pre-defined manner, failing to form a progressive semantic modeling process. To address these challenges, we propose KANResDiff to learn local residual diffusion with Kolmogorov-Arnold Network, thereby assigning distinct roles across stages for ambiguity modeling. Specifically, we propose Independent Time Encoding that offers spline-based time embeddings instead of linear ones from MLPs, which enhances the independence across inference stages and assigns progressive semantic roles to different stages. We propose Residual Schrodinger Bridge that injects deterministic residual prior with learnable weights by constructing local Schrodinger Bridge instead of following manually settings, achieving a flexible deterministic-stochastic interaction and stage-aware ambiguity modeling thanks to local optimal diffusion path. Extensive experimental results on two public datasets demonstrate that KANResDiff achieves SOTA performance on GED and HM-IoU, with maximum improvements of 16.8% and 7.7%, respectively, while maintaining competitive performance on the MDM metric. Source code is available at https://github.com/PerceptionComputingLab/KANResDiff.
Comments10 pages, 3 figures, MICCAI 2026 conference paper
