Classifying Nuclei Shape Heterogeneity in Breast Tumors with Skeletons

Abstract

In this study, we demonstrate the efficacy of scoring statistics derived from a medial axis transform, for differentiating tumor and non-tumor nuclei, in malignant breast tumor histopathology images. Characterizing nuclei shape is a crucial part of diagnosing breast tumors for human doctors, and these scoring metrics may be integrated into machine perception algorithms which aggregate nuclei information across a region to label whole breast lesions. In particular, we present a low-dimensional representation capturing characteristics of a skeleton extracted from nuclei. We show that this representation outperforms both prior morphological features, as well as CNN features, for classification of tumors. Nuclei and region scoring algorithms such as the one presented here can aid pathologists in the diagnosis of breast tumors.

Cite

Text

Falkenstein et al. "Classifying Nuclei Shape Heterogeneity in Breast Tumors with Skeletons." European Conference on Computer Vision Workshops, 2020. doi:10.1007/978-3-030-66415-2_20

Markdown

[Falkenstein et al. "Classifying Nuclei Shape Heterogeneity in Breast Tumors with Skeletons." European Conference on Computer Vision Workshops, 2020.](https://mlanthology.org/eccvw/2020/falkenstein2020eccvw-classifying/) doi:10.1007/978-3-030-66415-2_20

BibTeX

@inproceedings{falkenstein2020eccvw-classifying,
  title     = {{Classifying Nuclei Shape Heterogeneity in Breast Tumors with Skeletons}},
  author    = {Falkenstein, Brian and Kovashka, Adriana and Hwang, Seong Jae and Chennubhotla, S. Chakra},
  booktitle = {European Conference on Computer Vision Workshops},
  year      = {2020},
  pages     = {310-323},
  doi       = {10.1007/978-3-030-66415-2_20},
  url       = {https://mlanthology.org/eccvw/2020/falkenstein2020eccvw-classifying/}
}