Semi-Automatic Segmentation and Classification of Pap Smear Cells

Abstract

Cytologic screening has been widely used for detecting the cervical cancers. In this study, a semiautomatic PC-based cellular image analysis system was developed for segmenting nuclear and cytoplasmic contours and for computing morphometric and textual features to train support vector machine (SVM) classifiers to classify four different types of cells and to discriminate dysplastic from normal cells. A software program incorporating function, including image reviewing and standardized denomination of file names, was also designed to facilitate and standardize the workflow of cell analyses. Two experiments were conducted to verify the classification performance. The cross-validation results of the first experiment showed that average accuracies of 97.16% and 98.83%, respectively, for differentiating four different types of cells and in discriminating dysplastic from normal cells have been achieved using salient features (8 for four-cluster and 7 for two-cluster classifiers) selected with SVM recursive feature addition. In the second experiment, 70% (837) of the cell images were used for training and 30% (361) for testing, achieving an accuracy of 96.12% and 98.61% for four-cluster and two-cluster classifiers, respectively. The proposed system provides a feasible and effective tool in evaluating cytologic specimens.

A Semi - Automated Morphometric Assessment of Nuclei in Pap Smears Using Imagej

“…So, several attempts have been made to make these assessments more objective by using PC based morphometry. [15], [16], [17], [18] In the present study we have utilised ImageJ and three of its plugins to develop a simple macro to analyse nuclear parameters in normal and abnormal pap smears. In order to achieve this, we had to enhance the point of interest while removing the noise and suppressing the distracting background details.…”

Robust Nucleus/Cell Detection and Segmentation in Digital Pathology and Microscopy Images: A Comprehensive Review

“…Specifically, the algorithm can be summarized in four steps: 1) construct the graph and calculate the edge weight with an intensity similarity-based weighting function, 2) obtain seeded nodes with K labels, 3) compute the probability of each label in every node by solving a combinatorial Dirichlet problem, and 4) assign each node the label associated with the largest probability to obtain image segmentation. The random walk algorithm is applied to joint segmentation of nuclei and cytoplasm of Pap smear cells in [290], which consists of three major stages: 1) extract object edges with Sobel operator, 2) enhance the edges with a maximum gray-level gradient difference method, and 3) refine edges with random walk, which can separate overlapping objects. A fast random walker algorithm is presented in [291] for interactive blood smear cell segmentation, which improves the running time using offline precomputing.…”

Radial feature descriptors for cell classification and recommendation

“…We compare RFD with other 13 different feature sets, based on previously published work [6,7,8,9,10,11,12,13,14,15,16,17,18]; among them, 7 feature sets were designed specifically for cell classification [6,7,8,9,10,11,12].…”