Hierarchical Loss And Geometric Mask Refinement For Multilabel Ribs Segmentation
This work addresses a domain-specific problem in medical imaging for radiologists, with incremental improvements in segmentation quality.
The paper tackled automatic ribs segmentation and numeration in CT scans to improve assessment speed and reduce radiologist errors, achieving a new state-of-the-art 98.2% label accuracy on the RibSeg v2 dataset, surpassing the previous result by 6.7%.
Automatic ribs segmentation and numeration can increase computed tomography assessment speed and reduce radiologists mistakes. We introduce a model for multilabel ribs segmentation with hierarchical loss function, which enable to improve multilabel segmentation quality. Also we propose postprocessing technique to further increase labeling quality. Our model achieved new state-of-the-art 98.2% label accuracy on public RibSeg v2 dataset, surpassing previous result by 6.7%.