Segmentation and cell tracking of breast cancer cells

  • Authors:
  • Adele P. Peskin;Daniel J. Hoeppner;Christina H. Stuelten

  • Affiliations:
  • NIST, Boulder, CO;NIH, Bethesda, MD;NIH, Bethesda, MD

  • Venue:
  • ISVC'11 Proceedings of the 7th international conference on Advances in visual computing - Volume Part I
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

We describe a new technique to automatically segment and track the cell images of a breast cancer cell line in order to study cell migration and metastasis. Within each image observable cell characteristics vary widely, ranging from very bright completely bounded cells to barely visible cells with little to no apparent boundaries. A set of different segmentation algorithms are used in series to segment each cell type. Cell segmentation and cell tracking are done simultaneously, and no user selected parameters are needed. A new method for background subtraction is described and a new method of selective dilation is used to segment the barely visible cells. We show results for initial cell growth.