We propose the Topology-Preserving Segmentation Network, a deformation-based model that can extract objects in an image while maintaining their topological properties. This network generates segmentation masks that have the same topology as the template mask, even when trained with limited data. The network consists of two components: the Deformation Estimation Network, which produces a deformation map that warps the template mask to enclose the region of interest, and the Beltrami Adjustment Module, which ensures the bijectivity of the deformation map by truncating the associated Beltrami coefficient based on Quasiconformal theories. The proposed network can also be trained in an unsupervised manner, eliminating the need for labeled training data. This is achieved by incorporating an unsupervised segmentation loss. Our experimental results on various image datasets show that TPSN achieves better segmentation accuracy than state-of-the-art models with correct topology. Furthermore, we demonstrate TPSN's ability to handle multiple object segmentation.
TopoMortar: A dataset to evaluate image segmentation methods focused on topology accuracy
Department of Applied Mathematics and Computer Science, Technical University of Denmark(技术大学丹麦应用数学和计算机科学系);A.I. Virtanen Institute, University of Eastern Finland(东部芬兰大学A.I. Virtanen研究所);Department of Architecture, Faculty of Engineering, Sojo University, Japan(日本早稻田大学工学系建筑系)
Topology guaranteed segmentation of the human retina from OCT using convolutional neural networks
Portland State University(波特兰州立大学);Wilmer Eye Institute(威尔默眼科研究所);The Johns Hopkins University(约翰斯·霍普金斯大学);The Johns Hopkins University School of Medicine(约翰斯·霍普金斯大学医学院)
Disconnect to Connect: A Data Augmentation Method for Improving Topology Accuracy in Image Segmentation
DTU Compute, Technical University of Denmark(丹麦技术大学DTU计算学院);A.I. Virtanen Institute, University of Eastern Finland(东芬兰大学人工智能维尔塔宁研究所);Aarhus University(奥胡斯大学)
Topology-Preserving Image Segmentation with Spatial-Aware Persistent Feature Matching
Department of Electrical and Computer Engineering, University of California, San Diego(电气与计算机工程系,加州大学圣地亚哥分校);Department of Ophthalmology, University of California, San Diego(眼科学系,加州大学圣地亚哥分校)
TEDS-Net: Enforcing Diffeomorphisms in Spatial Transformers to Guarantee Topology Preservation in Segmentations
Institute of Biomedical Engineering(生物医学工程研究所);University of Oxford(牛津大学);Wellcome Centre for Integrative Neuroimaging(惠康综合神经成像中心);FMRIB(FMRIB(功能磁共振成像与脑研究中心));Australian Institute for Machine Learning (AIML)(澳大利亚机器学习研究所(AIML));University of Adelaide(阿德莱德大学);South Australian Health and Medical Research Institute (SAHMRI)(南澳大利亚健康与医学研究所(SAHMRI))
QIS : Interactive Segmentation via Quasi-Conformal Mappings
City University of Hong Kong(香港城市大学);Hong Kong Centre for Cerebro-Cardiovascular Health Engineering(香港心脑血管健康工程中心);Nankai University(南开大学);Chinese University of Hong Kong(香港中文大学)
Anatomically Conditioned Recurrent Refinement for Topology-Aware Circle of Willis Segmentation
Faculty of Electrical Engineering, Computer Science and Information Technology, Osijek, Croatia(奥西耶克大学电气工程、计算机科学与信息技术学院);Clinical Medical Center Osijek, Osijek, Croatia(奥西耶克临床医学中心);Ghent University, Dept. of Telecommunications and Information Processing, imec-TELIN-IPI, Ghent, Belgium(根特大学电信与信息处理系,imec-TELIN-IPI);Ghent University, Dept. of Telecommunications and Information Processing, TELIN-GAIM, Ghent, Belgium(根特大学电信与信息处理系,TELIN-GAIM)
DTEA: Dynamic Topology Weaving and Instability-Driven Entropic Attenuation for Medical Image Segmentation
School of Advanced Manufacturing, Guangdong University of Technology(广东技术大学先进制造学院);School of Electronic and Communication Engineering, Shenzhen Polytechnic University(深圳职业技术大学电子与通信工程学院);Guangzhou City University of Technology(广州市城市技术大学);Faculty of Science and Technology, University of Macau(澳门大学科学与技术学院);Department of Cardiology, Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences), Southern Medical University(广东省人民医院心内科(广东省医学科学院));Guangdong Cardiovascular Institute, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences(广东省心血管病研究所,广东省人民医院,广东省医学科学院)
Continuous-tone Simple Points: An $\ell_0$-Norm of Cyclic Gradient for Topology-Preserving Data-Driven Image Segmentation
Laboratory of Mathematics and Complex Systems (Ministry of Education), School of Mathematical Sciences, Beijing Normal University(数学与复杂系统实验室(教育部), 数学科学学院, 北京师范大学);State Key Laboratory of Remote Sensing Science, Faculty of Geographical Science, Beijing Normal University(遥感科学国家重点实验室, 地理科学学院, 北京师范大学)
Universal Topology Refinement for Medical Image Segmentation with Polynomial Feature Synthesis
Imperial College London(帝国理工学院);University of Oxford(牛津大学);Klinikum rechts der Isar, Technical University of Munich(慕尼黑工业大学右侧医院);Friedrich-Alexander-Universität Erlangen-Nürnberg(埃尔朗根-纽伦堡弗里德里希-亚历山大大学)
Topology Optimization in Medical Image Segmentation with Fast Euler Characteristic
Department of Computing, Imperial College London(帝国理工学院计算系);University of Oxford(牛津大学);Weill Cornell Medicine(韦尔医学院);Cornell Tech(康奈尔科技);TUM University Hospital, Technical University of Munich(慕尼黑技术大学医院);Friedrich-Alexander-Universität Erlangen-Nürnberg(埃尔兰根-纽伦堡弗里德里希-亚历山大大学);King’s College London(伦敦国王学院)
Topology-aware Mamba for Crack Segmentation in Structures
School of Computer Science and Engineering, Jiangsu University of Science and Technology(江苏科技大学计算机科学与工程学院);School of Electronic and Informatics Engineering, Jiangsu University(江苏大学电子与信息工程学院);School of Civil and Environmental Engineering, Oklahoma State University(俄克拉荷马州立大学土木与环境工程学院)
Preserving Cardiac Integrity: A Topology-Infused Approach to Whole Heart Segmentation
Imperial College London(帝国理工学院);Jilin university(吉林大学);National Heart and Lung Institute, Imperial College London(帝国理工学院国家心肺研究所);School of Biomedical Engineering & Imaging Sciences, King’s College London(伦敦国王学院生物医学工程与影像科学学院);Cardiovascular Research Centre, Royal Brompton Hospital(皇家布朗普顿医院心血管研究中心)
Costal Cartilage Segmentation with Topology Guided Deformable Mamba: Method and Benchmark
Plastic Surgery Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College(中国医学科学院北京协和医学院整形外科医院);School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳)理工学院);Nanjing University of Science and Technology(南京理工大学);The Second Hospital of Hebei Medical University(河北医科大学第二医院)
HyperTopo-Adapters: Geometry- and Topology-Aware Segmentation of Leaf Lesions on Frozen Encoders
Missouri Maize Computation and Vision Laboratory (MMCV)(密苏里州玉米计算与视觉实验室);Department of Electrical Engineering and Computer Science(电气工程与计算机科学系);University of Missouri(密苏里大学)
Topology-Aware Training and Spatial Diagnostics for Fiber Bundle Segmentation in Tracer Histology
Universidad Nacional Mayor de San Marcos(圣马科斯国立大学);Athinoula A. Martinos Center for Biomedical Imaging(阿西诺拉·A.马蒂诺斯生物医学成像中心);Massachusetts General Hospital(麻省总医院);Harvard Medical School(哈佛医学院);University of Rochester School of Medicine(罗切斯特大学医学院);McLean Hospital(麦克莱恩医院)
Chair for AI in Healthcare and Medicine, Technical University of Munich (TUM)(慕尼黑工业大学医疗保健与医学人工智能教席);TUM University Hospital(慕尼黑工业大学医院);School of Computation, Information and Technology, TUM(慕尼黑工业大学计算、信息与技术学院);Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心);Department of Computing, Imperial College London(伦敦帝国理工学院计算机系);Weill Cornell Medicine, Cornell University(康奈尔大学威尔医学院)