CELL PATTERN CLASSIFICATION OF INDIRECT IMMUNOFLUORESCENCE IMAGES

Abstract

AUTOIMMUNE DISEASES (ADS) DEVELOP WHEN THE IMMUNE SYSTEM OF THE BODY TREATS SOME HEALTHY CELLS AS `FOREIGNERS' AND ATTACKS THEM. ADS ARE AMONG THE TOP TEN LEADING CAUSES OF DEATH IN CHILDREN AND WOMEN IN ALL AGE GROUPS UP TO 64 YEARS. INDIRECT IMMUNOFLUORESCENCE (IIF) TEST IS USED TO CAPTURE HUMAN EPITHELIAL TYPE-2 (HEP-2) CELLS' IMAGES, WHERE THE DIFFERENT STAINING PATTERNS OF HEP-2 CELLS INDICATE THE STAGE AND TYPE OF THE AD. AUTOMATED CLASSIFICATION OF HEP-2 CELLS HAS ATTRACTED MUCH RESEARCH INTEREST IN RECENT YEARS. DESPITE THE EXTENSIVE RECENT WORK THAT HAS BEEN DONE IN THIS FIELD, THERE ARE STILL MANY CHALLENGES TO BE OVERCOME. THIS THESIS PRESENTS SOME EFFICIENT AND PRACTICAL METHODOLOGIES THAT OVERCOME THE CURRENT LIMITATIONS OF STATE-OF-THE-ART HEP-2 CELLS CLASSIFICATION METHODS. THE KEY CONTRIBUTIONS INCLUDE: AS THE FIRST STEP OF THE CELL IMAGES CLASSIFICATION APPROACHES, FEATURES ARE EXTRACTED FROM THE IMAGE PATCHES. BECAUSE OF THE SPARSE NATURE OF THE IMAGE PATCHES, A DI