Abnormal gastric cell segmentation based on shape using morphological operations

• Published in: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
• Main Author: Khalid N.E.A.; Samsudin N.; Hashim R.
• Format: Conference paper
• Language: English
• Published: 2012
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-84863883946&doi=10.1007%2f978-3-642-31075-1_54&partnerID=40&md5=0535521dd8af695f20a0bea89a157628

Cancer is the fourth leading cause of death among medically certified deaths in Malaysia. The most reliable diagnostic method to diagnose gastric adenocarcinoma is by inspecting the microscopic images of samples obtained through biopsy. These images are analyses by pathologist to identify the presence of cancer. However the process is time consuming and the interpretation varies with different pathologist. The application of image analysis techniques can assist pathologist towards a more efficient and faster diagnosis. Thus, this paper introduces an image analysis framework to automatically recognize and distinguished between normal gastric and gastric adenocarcinoma cells. The framework consist of the three phases of image analysis; preprocessing phase where the color tone issues are solved by component separation; processing phase which includes the thresholding and morphological techniques to segment the cells; post processing to identify the perimeter, area and roundness of the cells. This study shows that it is possible to automatically recognize and differentiate images with normal and abnormal cells. © 2012 Springer-Verlag.