A Conditional Generative Framework for Synthetic Data Augmentation in Segmenting Thin and Elongated Structures in Biological Images
This work addresses the data shortage problem for researchers in biological image analysis, offering an incremental improvement in synthetic data generation for filament segmentation.
The paper tackles the challenge of acquiring high-quality annotated datasets for segmenting thin and elongated filamentous structures in biological images by proposing a conditional generative framework based on Pix2Pix with a filament-aware structural loss to generate realistic synthetic images, and experiments show it outperforms models trained without synthetic data.
Thin and elongated filamentous structures, such as microtubules and actin filaments, often play important roles in biological systems. Segmenting these filaments in biological images is a fundamental step for quantitative analysis. Recent advances in deep learning have significantly improved the performance of filament segmentation. However, there is a big challenge in acquiring high quality pixel-level annotated dataset for filamentous structures, as the dense distribution and geometric properties of filaments making manual annotation extremely laborious and time-consuming. To address the data shortage problem, we propose a conditional generative framework based on the Pix2Pix architecture to generate realistic filaments in microscopy images from binary masks. We also propose a filament-aware structural loss to improve the structure similarity when generating synthetic images. Our experiments have demonstrated the effectiveness of our approach and outperformed existing model trained without synthetic data.