An Automated Method for Segmentation of Epithelial Cervical Cells in Images of ThinPrep
- 14 July 2009
- journal article
- Published by Springer Nature in Journal of Medical Systems
- Vol. 34 (6) , 1043-1058
Abstract
We present an automated method for segmentation of epithelial cells in images taken from ThinPrep scenes by a digital camera in a cytology lab. The method covers both steps of localization of cell objects in low resolution and detection of cytoplasm and nucleus boundary in high resolution. The underlying method makes use of geometric active contours as a powerful tool of segmentation. We also provide the analysis of the connected cells. For this purpose an automatic circular decomposition method is incorporated and adapted to the application by changing its segmentation condition. The results are evaluated numerically and compared with those of previous work in literature.Keywords
This publication has 21 references indexed in Scilit:
- Survey over image thresholding techniques and quantitative performance evaluationJournal of Electronic Imaging, 2004
- Adaptive document image binarizationPattern Recognition, 2000
- Unsupervised cell nucleus segmentation with active contoursSignal Processing, 1998
- Snakes, shapes, and gradient vector flowIEEE Transactions on Image Processing, 1998
- Optimal segmentation of cell imagesIEE Proceedings - Vision, Image, and Signal Processing, 1998
- A survey on evaluation methods for image segmentationPattern Recognition, 1996
- A Fast Level Set Method for Propagating InterfacesJournal of Computational Physics, 1995
- Shape modeling with front propagation: a level set approachIEEE Transactions on Pattern Analysis and Machine Intelligence, 1995
- Error measures for scene segmentationPattern Recognition, 1977
- Use of the Hough transformation to detect lines and curves in picturesCommunications of the ACM, 1972