Segmentation and classification of cervical cell images

Abstract

Cervical cancer can be prevented if it is detected and treated early. Pap smear test is a manual screening procedure used to detect cervical cancer and precancerous changes in an uterine cervix. However, this procedure is costly and it may result in inaccurate diagnoses due to human error like intra- and interobserver variability. Therefore, a computer-assisted screening system will be very bene cial to prevent cervical cancer if it increases the reliability of diagnoses. In this thesis, we propose a computer-assisted diagnosis system which helps cyto-technicians by sorting cells in a Pap smear slide according to their abnormality degree. There are three main components of such a system. Firstly, cells along with their nuclei are located using a segmentation procedure on an image taken using a microscope. Then, features describing these segmented cells are extracted. Finally, the cells are sorted according to their abnormality degree based on the extracted features. Di erent from the related studies that require images of a single cervical cell, we propose a non-parametric generic segmentation algorithm that can also handle images of overlapping cells. We use thresholding as the rst phase to extract background regions for obtaining remaining cell regions. The second phase consists of segmenting the cell regions by a non-parametric hierarchical segmentation algorithm that uses the spectral and shape information as well as the gradient information. The last phase aims to partition the cell region into true structures of each nucleus and the whole cytoplasm area by classifying the nal segments as nucleus or cytoplasm region. We evaluate our segmentation method both quantitatively and qualitatively using two data sets.By proposing an unsupervised screening system, we aim to approach the problem in a di erent way when compared to the related studies that concentrate on classi cation. In order to rank the cells in a Pap slide, we rst perform hierarchical clustering on 14 di erent cell features. The initial ordering of the cells is determined as the leaf ordering of the constructed hierarchical tree. Then, this initial ordering is improved by applying an optimal leaf ordering algorithm. The experiments with ground truth data show the e ectiveness of the proposed approach under di erent experimental settings.