Multi-channels statistical and morphological features based mitosis detection in breast cancer histopathology

Humayun Irshad, Ludovic Roux and Daniel Racoceanu

Working Paper from Harvard University OpenScholar

Abstract: Accurate counting of mitosis in breast cancer histopathology plays a critical role in the grading process. Manual counting of mitosis is tedious and subject to considerable inter- and intra-reader variations. This work aims at improving the accuracy of mitosis detection by selecting the color channels that better capture the statistical and morphological features having mitosis discrimination from other objects. The proposed framework includes comprehensive analysis of first and second order statistical features together with morphological features in selected color channels and a study on balancing the skewed dataset using SMOTE method for increasing the predictive accuracy of mitosis classification. The proposed framework has been evaluated on MITOS data set during an ICPR 2012 contest and ranked second from 17 finalists. The proposed framework achieved 74% detection rate, 70% precision and 72% F-Measure. In future work, we plan to apply our mitosis detection tool to images produced by different types of slide scanners, including multi-spectral and multi-focal microscopy.

Pages: 6091?6094
Date: 2014-12
References: Add references at CitEc
Citations:

Downloads: (external link)
http://scholar.harvard.edu/humayun/node/221791

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:qsh:wpaper:221791

Access Statistics for this paper

More papers in Working Paper from Harvard University OpenScholar Contact information at EDIRC.
Bibliographic data for series maintained by Richard Brandon ( this e-mail address is bad, please contact ).