Segmentation of Tubular Structures Using Iterative Training with Tailored Samples
This work addresses the discrepancy between training and inference samples in CNN-based methods for tubular structure segmentation, which is incremental but improves performance in applications like satellite and medical imaging.
The authors tackled the problem of segmenting tubular structures by introducing an iterative training scheme that generates tailored samples for minimal path methods, achieving state-of-the-art results on three public datasets for both segmentation masks and centerlines.
We propose a minimal path method to simultaneously compute segmentation masks and extract centerlines of tubular structures with line-topology. Minimal path methods are commonly used for the segmentation of tubular structures in a wide variety of applications. Recent methods use features extracted by CNNs, and often outperform methods using hand-tuned features. However, for CNN-based methods, the samples used for training may be generated inappropriately, so that they can be very different from samples encountered during inference. We approach this discrepancy by introducing a novel iterative training scheme, which enables generating better training samples specifically tailored for the minimal path methods without changing existing annotations. In our method, segmentation masks and centerlines are not determined after one another by post-processing, but obtained using the same steps. Our method requires only very few annotated training images. Comparison with seven previous approaches on three public datasets, including satellite images and medical images, shows that our method achieves state-of-the-art results both for segmentation masks and centerlines.