HyperAIHyperAI
2 months ago

Circle Representation for Medical Object Detection

Nguyen, Ethan H. ; Yang, Haichun ; Deng, Ruining ; Lu, Yuzhe ; Zhu, Zheyu ; Roland, Joseph T. ; Lu, Le ; Landman, Bennett A. ; Fogo, Agnes B. ; Huo, Yuankai
Circle Representation for Medical Object Detection
Abstract

Box representation has been extensively used for object detection in computervision. Such representation is efficacious but not necessarily optimized forbiomedical objects (e.g., glomeruli), which play an essential role in renalpathology. In this paper, we propose a simple circle representation for medicalobject detection and introduce CircleNet, an anchor-free detection framework.Compared with the conventional bounding box representation, the proposedbounding circle representation innovates in three-fold: (1) it is optimized forball-shaped biomedical objects; (2) The circle representation reduced thedegree of freedom compared with box representation; (3) It is naturally morerotation invariant. When detecting glomeruli and nuclei on pathological images,the proposed circle representation achieved superior detection performance andbe more rotation-invariant, compared with the bounding box. The code has beenmade publicly available: https://github.com/hrlblab/CircleNet

Circle Representation for Medical Object Detection | Latest Papers | HyperAI