A fast and efficient segmentation scheme for cell microscopic image.

Microscopic cellular image segmentation schemes must be efficient for reliable analysis and fast to process huge quantity of images. Recent studies have focused on improving segmentation quality. Several segmentation schemes have good quality but processing time is too expensive to deal with a great number of images per day. For segmentation schemes based on pixel classification, the classifier design is crucial since it is the one which requires most of the processing time necessary to segment an image. The main contribution of this work is focused on how to reduce the complexity of decision functions produced by support vector machines (SVM) while preserving recognition rate. Vector quantization is used in order to reduce the inherent redundancy present in huge pixel databases (i.e. images with expert pixel segmentation). Hybrid color space design is also used in order to improve data set size reduction rate and recognition rate. A new decision function quality criterion is defined to select good trade-off between recognition rate and processing time of pixel decision function. The first results of this study show that fast and efficient pixel classification with SVM is possible. Moreover posterior class pixel probability estimation is easy to compute with Platt method. Then a new segmentation scheme using probabilistic pixel classification has been developed. This one has several free parameters and an automatic selection must dealt with, but criteria for evaluate segmentation quality are not well adapted for cell segmentation, especially when comparison with expert pixel segmentation must be achieved. Another important contribution in this paper is the definition of a new quality criterion for evaluation of cell segmentation. The results presented here show that the selection of free parameters of the segmentation scheme by optimisation of the new quality cell segmentation criterion produces efficient cell segmentation.

A neural network architecture for data classification.

This article aims at showing an architecture of neural networks designed for the classification of data distributed among a high number of classes. A significant gain in the global classification rate can be obtained by using our architecture. This latter is based on a set of several little neural networks, each one discriminating only two classes. The specialization of each neural network simplifies their structure and improves the classification. Moreover, the learning step automatically determines the number of hidden neurons. The discussion is illustrated by tests on databases from the UCI machine learning database repository. The experimental results show that this architecture can achieve a faster learning, simpler neural networks and an improved performance in classification.

Segmentation of color images from serous cytology for automated cell classification.

OBJECTIVE: To design an automated system for the classification of cells based on analysis of serous cytology, with the aim of segmenting both cytoplasm and nucleus using color information from the images as the main characteristic of the cells. STUDY DESIGN: The segmentation strategy uses color information coupled with mathematical morphology tools, such as watersheds. Cytoplasm and nuclei of all diagnostic cells are retained; erythrocytes and debris are eliminated. Special techniques are used for the separation of clustered cells. RESULTS: A large set of cells was assessed by experts to score the segmentation success rate. All cells were segmented whatever their spatial configurations. The average success rate was 92.5% for nuclei and 91.1% for cytoplasm. CONCLUSION: This color information-based segmentation of images of serous cells is accurate and provides a useful tool. This segmentation strategy will improve the automated classification of cells.