Gastric Adenocarcinoma Cell Microscopic Image Analysis And Processing System

Computer-aided image processing using medical image characteristic is an important application in the field of medical image processing and analysis. The segmentation and recognition methods of cell images have developed rapidly with the development of computer technology and become the one-up research task in present image domain. It is absolutely helpful for clinical diagnosis and medical studies that computer-aided diagnosis system of cancerous tissues is founded, especially in the case of lack of specialists. In this paper, the research of the processing and analysis system of gastric adenocarcinoma image and the knowledge base of gastric adenocarcinoma cell characteristics and some image segmentation methods of this cell image are focused on.The pathology experts give the diagnosis all by the knowledge base of gastric adenocarcinoma cell characteristics. And this information provides the foundation of theory for researching of the processing and analysis system of gastric adenocarcinoma image. Importantly it's also the linchpin of evaluating the result of the segmentation.So firstly,some characteristics of gastric adenocarcinoma cell are analyzed. There are six parameters of veins, three color character parameters and six morphological parameters, and then the methods of measuring them are presented.In order to found the knowledge base of gastric adenocarcinoma cell characteristics and segment the nuclei of every cell and extract the features,the technology of pre-processing for the gastric adenocarcinoma cell image and the segmentation technique of getting the exact region of cell nuclei have been researched,especially for the overlapping parts. The paper presents an improved adaptive threshold algorithm based on 2D Otsu to get the accurate region of cell nuclei and an estimate algorithm of separate line based on marker-controlled watershed segmentation to isolate the overlapped cells. The methods above get the contour of single cell and extract the morphometry, heterochromatic and textual features correctly, which is the base of automatic recognition of gastric adenocarcinoma cells.And after that , the object-oriented and modular design pattern are adopted in the work.The system we proposed has six basic parts, including image acquiring module, basic image processing module, medical image processing module, metadata management module of medical image,image retrieval module and recognition module. Test results show that doctors improve their efficiency and accuracy aided by our system.