Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Diffusion Model-Based Data Augmentation for Enhanced Neuron Segmentation

Liuyun Jiang, Yanchao Zhang, Jinyue Guo, Yizhuo Lu, Ruining Zhou, Hua Han

TL;DR

This work tackles the data scarcity problem in neuron segmentation from electron microscopy by introducing a diffusion-based data augmentation framework. It combines a 3D EM image synthesis module with a biology-guided mask remodeling component to generate diverse, structurally plausible image–label pairs, enabling voxel-level control via a resolution-aware conditional diffusion model that leverages MSC and RPGE with a Mamba backbone. The approach yields substantial improvements in segmentation accuracy under low-annotation regimes, achieving notable ARAND gains and consistent benefits across multiple segmentation backbones, while also delivering improved 3D-FID scores for generated images. The method advances practical neuron reconstruction by providing high-fidelity synthetic data while respecting biological structure, and it is made reproducible with publicly available code.

Abstract

Neuron segmentation in electron microscopy (EM) aims to reconstruct the complete neuronal connectome; however, current deep learning-based methods are limited by their reliance on large-scale training data and extensive, time-consuming manual annotations. Traditional methods augment the training set through geometric and photometric transformations; however, the generated samples remain highly correlated with the original images and lack structural diversity. To address this limitation, we propose a diffusion-based data augmentation framework capable of generating diverse and structurally plausible image-label pairs for neuron segmentation. Specifically, the framework employs a resolution-aware conditional diffusion model with multi-scale conditioning and EM resolution priors to enable voxel-level image synthesis from 3D masks. It further incorporates a biology-guided mask remodeling module that produces augmented masks with enhanced structural realism. Together, these components effectively enrich the training set and improve segmentation performance. On the AC3 and AC4 datasets under low-annotation regimes, our method improves the ARAND metric by 32.1% and 30.7%, respectively, when combined with two different post-processing methods. Our code is available at https://github.com/HeadLiuYun/NeuroDiff.

Diffusion Model-Based Data Augmentation for Enhanced Neuron Segmentation

TL;DR

This work tackles the data scarcity problem in neuron segmentation from electron microscopy by introducing a diffusion-based data augmentation framework. It combines a 3D EM image synthesis module with a biology-guided mask remodeling component to generate diverse, structurally plausible image–label pairs, enabling voxel-level control via a resolution-aware conditional diffusion model that leverages MSC and RPGE with a Mamba backbone. The approach yields substantial improvements in segmentation accuracy under low-annotation regimes, achieving notable ARAND gains and consistent benefits across multiple segmentation backbones, while also delivering improved 3D-FID scores for generated images. The method advances practical neuron reconstruction by providing high-fidelity synthetic data while respecting biological structure, and it is made reproducible with publicly available code.

Abstract

Neuron segmentation in electron microscopy (EM) aims to reconstruct the complete neuronal connectome; however, current deep learning-based methods are limited by their reliance on large-scale training data and extensive, time-consuming manual annotations. Traditional methods augment the training set through geometric and photometric transformations; however, the generated samples remain highly correlated with the original images and lack structural diversity. To address this limitation, we propose a diffusion-based data augmentation framework capable of generating diverse and structurally plausible image-label pairs for neuron segmentation. Specifically, the framework employs a resolution-aware conditional diffusion model with multi-scale conditioning and EM resolution priors to enable voxel-level image synthesis from 3D masks. It further incorporates a biology-guided mask remodeling module that produces augmented masks with enhanced structural realism. Together, these components effectively enrich the training set and improve segmentation performance. On the AC3 and AC4 datasets under low-annotation regimes, our method improves the ARAND metric by 32.1% and 30.7%, respectively, when combined with two different post-processing methods. Our code is available at https://github.com/HeadLiuYun/NeuroDiff.
Paper Structure (12 sections, 3 equations, 4 figures, 5 tables)

This paper contains 12 sections, 3 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Resolution-Aware Conditional Diffusion
  4. Biology-Guided Mask Remodeling
  5. Experiments
  6. Datasets and Metrics
  7. Implementation Details
  8. Comparison with State-of-the-art Methods
  9. Ablation Studies and Analysis
  10. Conclusion
  11. Compliance with ethical standards
  12. Acknowledgements

Figures (4)

  • Figure 1: (a) Illustration of the neuron reconstruction paradigm: EM images are processed by a segmentation network to generate affinity maps, which are aggregated by a post-processing algorithm to obtain the final reconstruction. (b) Traditional data acquisition relies on manual annotation for limited-scale training data. (c) Our method generates new image–label pairs to enrich the training dataset.
  • Figure 2: Architecture of the proposed Resolution-Aware Conditional Diffusion model.
  • Figure 3: Illustration of the Biology-Guided Mask Remodeling.
  • Figure 4: Visualization of generated image quality. Blue arrows highlight potential artifacts, while red arrows indicate possible membrane loss.