Neural classification of cytological smears from the cervix

Background: Cytological smears obtained from the cervix are routinely exami ned under the microscope as part of screening programs for the early detect ion of cervical cancer. The aim of the present study was to investigate whe ther a simple feature extraction approach using only standard image process ing techniques combined with a neural classifier would lead to acceptable r esults that might serve as a starting point for the development of a fully automated screening system. Materials and methods: Gray-value images of 106 cervical smears (512 x 512 pixels) divided into two groups - inconspicuous (57) and atypical (49) - by an experienced pathologist on the basis of the original smears were employ ed to evaluate the method. From these images, 31 features quantifying prope rties of either the cell nucleus or the cytoplasm were extracted. These fea tures were categorized with three different architectures of a neural class ifier: learning vector quantization (LVQ), multilayer perceptron (MLP) and a single perceptron. Conclusions: The results show a reclassification accuracy of: about 91% for all three algorithms. Sensitivity was uniform at approximately 78 %, and s pecificity varied between 75 % and 91 % in the leave-one-out evaluation. Th ese very good results provide strong encouragement for further studies invo lving PAP scores and colour images.