Automated
Nuclei Segmentation Approach based on Mathematical Morphology for Cancer
Scoring in Breast Tissue Images
Aymen Mouelhi1, Mounir Sayadi1, Farhat
Fnaiech1 and Karima Mrad2
1Laboratory of
Signal Image and Energy Mastery, ENSIT-University of Tunis, Tunisia.
2Morbid Anatomy
Service, Salah Azaiez Institute of Oncology, Tunisia
Abstract: In
this work, we propose an automated approach able to perform accurate nuclear
segmentation in immunohistochemical breast tissue images in order to provide
quantitative evaluation of estrogen or progesterone receptor status that will
help pathologists in their diagnosis. The presented method is based on color
deconvolution and an enhanced morphological processing, which is used to
identify positive stained nuclei and to separate all clustered nuclei in the
microscopic image for a subsequent cancer scoring. Experiments on several
breast cancer images of different patients admitted into the Tunisian Salah
Azaiez Cancer Center, show the efficiency of the proposed method when compared
to the manual evaluation of experts. On the whole image database, we recorded
more than 97% for both accuracy of detected nuclei and cancer scoring over the
truths provided by experienced pathologists.
Keywords: Breast
cancer, immunohistochemical image analysis, color deconvolution, morphological
operators.
Received September 11,
2014; accepted March 23, 2015