The Research and Exploration of Soft X-Ray Microscopy on the Image Identification System of Squamocellular Cancer of Esophagus

Abstract

Nowadays, soft X-Ray microscopy technology is one of the hottest methods that researches and observes the biological samples of natural state, whose resolution is higher than the imaging drawing of optical microscope and near to the imaging drawing of electron microscope. The technology is simple for the preparation of biological sample, which needs only ultramicrocut. And the working procedure of traditional pathology check, such as freezing, sealing wax, dehydration and dyeing, etc. isn't needed. A kind of image recognition system is selected in the paper, which identifies and researches the tissue morphological characteristics of ultrastructure imaging drawing of squamocellular cancer of esophagus obtained by soft X-Ray microscopy. The system mainly consists of components of image preprocessing, image segmentation, cellular feature extraction and cellular feature recognition, etc. Image preprocessing includes grayscale transformation, histogram adjustment, etc. Image segmentation includes segmentation based on the threshold, image processing based on the structure morphology of cell tissue, and edge detection, etc. In cellular feature extraction, the method of extracting area connected is used. And in cellular feature recognition, the judgment of area ratio of nucleus and cytoplasm is adopted. According to the morphology characteristics of tissue ultrastructure of squamocellular cancer of esophagus, a kind of cancer identification method based on the area threshold is employed emphatically in the paper, aiming at distinguishing the normal cell and cancer cell successfully. The technology application on the aspects of clinical diagnosis and differentiate diagnosis of the image recognition system in the paper still needs to be researched further.