What is it about?
Segmentation of airways in CT lung images is crucial for medical diagnosis. Manual segmentation being time expensive, automatic airway segmentation based on U-Net with Attention mechanism is being proposed in this work. The model has been validated using VESSEL12 and EXACT09 datasets with DSC scores of 95.21% and 95.80% respectively.
Featured Image
Why is it important?
This is a novel work of airway segmentation based on U-Net architecture employing the Attention mechanism for extracting complex and multi-sized airways from lung images, thus increasing the efficiency of U-Net model.
Perspectives
As not much work has been done in the area of airway segmentation because of the difficulty of generating ground truth, this article is a contribution towards the diagnosis of pathologies in lung parenchyma and airway structures. it will open doors for people to continue working in this area.
anita khanna
Read the Original
This page is a summary of: A Deep Attention-based U-Net for Airways Segmentation in Computed
Tomography Images, Current Medical Imaging Formerly Current Medical Imaging Reviews, April 2023, Bentham Science Publishers,
DOI: 10.2174/1573405618666220630151409.
You can read the full text:
Contributors
The following have contributed to this page
