Preserving instance continuity and length in segmentation through connectivity-aware loss computation

2025-09-03

AI Summary

Biomedical image segmentation of elongated structures—such as axon initial segments—is prone to discontinuities due to signal dropout, leading to inaccurate instance-length measurements. To address this, we propose a topology-aware segmentation method that explicitly enforces structural continuity. Our approach introduces two novel loss functions: a negative centerline loss to suppress centerline fragmentation and a simplified topological loss to constrain connectivity. These are integrated with downsampling adaptation and voxel-spacing correction to incorporate structural priors directly into training. Implemented within a CNN framework, the method performs connectivity-aware mask optimization. Evaluated on a 3D light-sheet fluorescence microscopy dataset, our method significantly reduces segmentation discontinuities—especially in low signal-to-noise ratio regions—and achieves an average 32.7% reduction in downstream instance-length measurement error. This enhances both the accuracy and robustness of quantitative morphological analysis.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Structured Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Security and Privacy: Data transparency and provenance

Abstract

In many biomedical segmentation tasks, the preservation of elongated structure continuity and length is more important than voxel-wise accuracy. We propose two novel loss functions, Negative Centerline Loss and Simplified Topology Loss, that, applied to Convolutional Neural Networks (CNNs), help preserve connectivity of output instances. Moreover, we discuss characteristics of experiment design, such as downscaling and spacing correction, that help obtain continuous segmentation masks. We evaluate our approach on a 3D light-sheet fluorescence microscopy dataset of axon initial segments (AIS), a task prone to discontinuity due to signal dropout. Compared to standard CNNs and existing topology-aware losses, our methods reduce the number of segmentation discontinuities per instance, particularly in regions with missing input signal, resulting in improved instance length calculation in downstream applications. Our findings demonstrate that structural priors embedded in the loss design can significantly enhance the reliability of segmentation for biological applications.

Problem

• Preserving connectivity in biomedical segmentation tasks
• Reducing discontinuities in elongated structure segmentation
• Improving instance length calculation accuracy

Innovation

• Connectivity-aware loss functions for segmentation continuity
• Negative Centerline Loss and Simplified Topology Loss
• Downscaling and spacing correction design improvements