TopoMortar: A dataset to evaluate image segmentation methods focused on topology accuracy

作     者：Valverde, Juan Miguel Koga, Motoya Otsuka, Nijihiko Dahl, Anders Bjorholm 

作者机构：Department of Applied Mathematics and Computer Science Technical University of Denmark Denmark A.I. Virtanen Institute University of Eastern Finland Finland Department of Architecture Faculty of Engineering Sojo University Japan 

出 版 物：《arXiv》 (arXiv)

年 卷 期：2025年

核心收录：

主　　题：Musculoskeletal system 

摘      要：We present TopoMortar, a brick wall dataset that is the first dataset specifically designed to evaluate topology-focused image segmentation methods, such as topology loss functions. TopoMortar enables to investigate in two ways whether methods incorporate prior topological knowledge. First, by eliminating challenges seen in real-world data, such as small training set, noisy labels, and out-of-distribution test-set images, that, as we show, impact the effectiveness of topology loss functions. Second, by allowing to assess in the same dataset (TopoMortar) topology accuracy across dataset challenges, isolating dataset-related effects from the effect of incorporating prior topological knowledge. In these two experiments, it is deliberately difficult to improve topology accuracy without actually leveraging topology information, thus, permitting to attribute an improvement in topology accuracy to the incorporation of prior topological knowledge. To this end, TopoMortar includes three types of labels (accurate, pseudo-labels, noisy labels), two fixed training sets (large and small), and in-distribution and out-of-distribution test-set images. We compared eight loss functions on TopoMortar, and we found that clDice generally achieved the most topologically accurate segmentations, Skeleton Recall loss performed best particularly with noisy labels, and the relative advantageousness of the other loss functions depended on the experimental setting. Additionally, we show that simple methods, such as data augmentation and self-distillation, can elevate Cross entropy Dice loss to surpass most topology loss functions, and that those simple methods can enhance topology loss functions as well. clDice and Skeleton Recall loss, both skeletonization-based loss functions, were also the fastest to train, making this type of loss function a promising research direction. TopoMortar and our code can be found at https://***/jmlipman/TopoMortar. © 2025, CC BY.