Hierarchical segmentation of surgical scenes in laparoscopy

Pritesh Mehta, David Owen, Maria Grammatikopoulou, Lucy Culshaw, Karen Kerr, Danail Stoyanov, Imanol Luengo

Abstract: Segmentation of surgical scenes may provide valuable information for real-time guidance and post-operative analysis. However, in some surgical video frames there is unavoidable ambiguity, leading to incorrect predictions of class or missed detections. In this work, we propose a novel method that alleviates this problem by introducing a hierarchy and associated hierarchical inference scheme that allows broad anatomical structures to be predicted when fine-grained structures cannot be reliably distinguished.