{
  "id": "2002.01020",
  "version": "v1",
  "published": "2020-02-03T21:20:50.000Z",
  "updated": "2020-02-03T21:20:50.000Z",
  "title": "Bending Loss Regularized Network for Nuclei Segmentation in Histopathology Images",
  "authors": [
    "Haotian Wang",
    "Min Xian",
    "Aleksandar Vakanski"
  ],
  "comment": "4 pages, 5 figures, 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI), accepted",
  "categories": [
    "eess.IV",
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "Separating overlapped nuclei is a major challenge in histopathology image analysis. Recently published approaches have achieved promising overall performance on public datasets; however, their performance in segmenting overlapped nuclei are limited. To address the issue, we propose the bending loss regularized network for nuclei segmentation. The proposed bending loss defines high penalties to contour points with large curvatures, and applies small penalties to contour points with small curvature. Minimizing the bending loss can avoid generating contours that encompass multiple nuclei. The proposed approach is validated on the MoNuSeg dataset using five quantitative metrics. It outperforms six state-of-the-art approaches on the following metrics: Aggregate Jaccard Index, Dice, Recognition Quality, and Pan-optic Quality.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-02-03T21:20:50.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "bending loss regularized network",
      "nuclei segmentation",
      "bending loss defines high penalties",
      "contour points",
      "applies small penalties"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 4,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}