Segmentation of candidate bacillus objects in images of Ziehl-Neelsen-stained sputum smears using deformable models

Automated microscopy for the detection of tuberculosis (TB) in sputum smears seeks to address the strain on technicians and to achieve faster diagnosis in order to cope with the rising number of TB cases. Image processing techniques provide a useful alternative to the conventional, manual analysis of sputum smears for diagnosis. In the project described here, the use of parametric and geometric deformable models was explored for segmentation of TB bacilli in images of Ziehl-Neelsen-stained sputum smears for automated TB diagnosis. The goal of segmentation is to produce candidate bacillus objects for input into a classifier.

Grantor dc:publisher.institution

Division of Biomedical Engineering

Year dc:date.issued

2009

Author dc:creator

• Dendere, Ronald

Advisor dc:contributor.advisor

• Douglas, Tania S

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/3232

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/3232