Autofocusing and image segmentation in microscopy for automatic detection of tuberculosis in sputum smears

• Main Author: Russell, Megan J
• Other Authors: Douglas, Tania S
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2014
• Subjects: Medicine
• Online Access: http://hdl.handle.net/11427/3198

Includes bibliographical references (leaves 89-92). === The World Health Organisation (WHO) has declared that tuberculosis (TB) is a global emergency. With the common combination of TB and AIDS in sub-Saharan Africa, the number of people dying from TB has increased enormously (WHO, 1996). The smart microscope that is under development in the MRC/UCT Medical Imaging Research Unit would considerably aid the fast detection and tracking of TB. The microscope will automatically detect the presence of TB bacilli in Ziehl-Neelsen (ZN) stained sputum smears. The microscope consists of a digital camera, motorised stages, and will have an on-board processor containing the algorithms for autofocusing the microscope and for automatically identifying TB bacteria on the slides. This thesis deals with autofocusing the microscope and segmenting TB bacilli from focused sputum smear images.