Topology-Preserving Deep Image Segmentation

Abstract

Segmentation algorithms are prone to make topological errors on fine-scale structures, e.g., broken connections. We propose a novel method that learns to segment with correct topology. In particular, we design a continuous-valued loss function that enforces a segmentation to have the same topology as the ground truth, i.e., having the same Betti number. The proposed topology-preserving loss function is differentiable and we incorporate it into end-to-end training of a deep neural network. Our method achieves much better performance on the Betti number error, which directly accounts for the topological correctness. It also performs superiorly on other topology-relevant metrics, e.g., the Adjusted Rand Index and the Variation of Information. We illustrate the effectiveness of the proposed method on a broad spectrum of natural and biomedical datasets.

11 pages, 7 figures

References in corpus (3)

• Rethinking Atrous Convolution for Semantic Image Segmentation
• Sliced Wasserstein Kernel for Persistence Diagrams
• A Topological Regularizer for Classifiers via Persistent Homology

Cited by in corpus (26)

• A Topological Loss Function for Deep-Learning based Image Segmentation using Persistent Homology
• Uncovering the Topology of Time-Varying fMRI Data using Cubical Persistence
• SemiCurv: Semi-Supervised Curvilinear Structure Segmentation
• Topological Learning for Brain Networks
• clDice -- A Novel Topology-Preserving Loss Function for Tubular Structure Segmentation
• Path homologies of deep feedforward networks
• Topology-preserving Scan-based Immersed Isogeometric Analysis
• TopoTxR: A topology-guided deep convolutional network for breast parenchyma learning on DCE-MRIs
• A plug-and-play framework for curvilinear structure segmentation based on a learned reconnecting regularization
• Some open questions on morphological operators and representations in the deep learning era
• Trigger Hunting with a Topological Prior for Trojan Detection
• Fuzzy c-Means Clustering for Persistence Diagrams
• Topological Information Retrieval with Dilation-Invariant Bottleneck Comparative Measures
• Fast Topological Clustering with Wasserstein Distance
• Image Segmentation with Topological Priors
• Combining Geometric and Topological Information for Boundary Estimation
• Multi-task learning for joint weakly-supervised segmentation and aortic arch anomaly classification in fetal cardiac MRI
• Bayesian SegNet for Semantic Segmentation with Improved Interpretation of Microstructural Evolution During Irradiation of Materials
• Quantifying Topology In Pancreatic Tubular Networks From Live Imaging 3D Microscopy
• TopoTxR: A Topological Biomarker for Predicting Treatment Response in Breast Cancer
• Topological Node2vec: Enhanced Graph Embedding via Persistent Homology
• Deep Neural Network for 3D Surface Segmentation based on Contour Tree Hierarchy
• Stateless actor-critic for instance segmentation with high-level priors
• High-level Prior-based Loss Functions for Medical Image Segmentation: A Survey
• Abstraction, Reasoning and Deep Learning: A Study of the "Look and Say" Sequence
• Topological Regularization for Dense Prediction