Using hyperchromasia to classify colonic microscopic images

• Main Authors: Gan Lim, Laurence A., Naguib, Raouf N. G., Dadios, Elmer P., Avila, Jose Maria C.
• Format: text
• Published: Animo Repository 2022
• Subjects: Colon (Anatomy)—Cancer—Imaging; Neural networks (Computer science); Computer Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/5844
• Institution: De La Salle University

In this paper, the effect of nuclear hyperchromasia on cancerous histopathological images is presented as a discriminating feature in the classification of colon microscopic images. All the images that were used in the study were pre-classified by a human expert into three (3) categories or cases: normal, adenomatous polyp, and adenocarcinoma or cancerous. The images were selected from a pool of colon images in such a way that there were 25 images in each category, totalling 75 images in all. Of the 25 images in each case, 15 were devoted for training of the algorithm while 10 were used for testing or actual classification. Success and failure in the classification are presented and summarized in a confusion matrix.