{"id":{"repo_id":"cape-town","oai_identifier":"oai:open.uct.ac.za:11427/3271"},"canonical_url":"https://search.dev.ndltd.org/etd/cape-town/oai:open.uct.ac.za:11427/3271","repository":{"repo_id":"cape-town","name":"University of Cape Town","base_url":"https://open.uct.ac.za/oai/request"},"display":{"title":"Computer-aided diagnosis of tuberculosis in paediatric chest X-rays using local textural analysis","abstract":"This report presents a computerised tool to analyse the appearance of the lung fields in paediatric chest X-rays to detect the presence of tuberculosis. The computer aided diagnosis (CAD) tool consists of 4 phases: 1) lung field segmentation; 2) lung field subdivision; 3) feature extraction and 4) classification. Lung field segmentation is performed using a semi-automatic implementation of the active shape model algorithm. Two approaches to subdividing the lung fields into regions of interest are compared. The first divides each lung field into 21 overlapping regions of varying sizes, resulting in a total of 42 regions per image; this approach is called the big region approach. The second approach divides the lung fields into a large number of overlapping circular regions of interest. The circular regions have a radius of 32 pixels and are placed on an 8 x 8 pixel grid. This approach is called the circular region approach. Textural features are extracted from each of the regions using the moments of responses to a multiscale bank of Gaussian filters. Additional positional features are added to the circular regions.","abstract_html":"This report presents a computerised tool to analyse the appearance of the lung fields in paediatric chest X-rays to detect the presence of tuberculosis. The computer aided diagnosis (CAD) tool consists of 4 phases: 1) lung field segmentation; 2) lung field subdivision; 3) feature extraction and 4) classification. Lung field segmentation is performed using a semi-automatic implementation of the active shape model algorithm. Two approaches to subdividing the lung fields into regions of interest are compared. The first divides each lung field into 21 overlapping regions of varying sizes, resulting in a total of 42 regions per image; this approach is called the big region approach. The second approach divides the lung fields into a large number of overlapping circular regions of interest. The circular regions have a radius of 32 pixels and are placed on an 8 x 8 pixel grid. This approach is called the circular region approach. Textural features are extracted from each of the regions using the moments of responses to a multiscale bank of Gaussian filters. Additional positional features are added to the circular regions.","abstract_has_math":false,"creators":["Mouton, Andre"],"institution":"Department of Human Biology","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Douglas, Tania S"],"committee_chairs":[],"committee_members":[],"year":2009,"date_issued":"2009","date_published":"2009","updated_at":"2026-07-22T22:22:57Z","subjects":[],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/11427/3271","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Douglas, Tania S"]},{"key":"dc:creator","label":"Author","values":["Mouton, Andre"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2014-07-28T18:17:35Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2014-07-28T18:17:35Z"]},{"key":"dc:date.issued","label":"Date","values":["2009"]},{"key":"dc:publisher.department","label":"Dc Publisher Department","values":["Department of Human Biology"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Cape Town"]},{"key":"dc:type","label":"Dc Type","values":["Master Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["Masters"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["MSc"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://hdl.handle.net/11427/3271"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Includes abstract.","Includes bibliographical references (leaves 99-103)."]},{"key":"dc:description.abstract","label":"Abstract","values":["This report presents a computerised tool to analyse the appearance of the lung fields in paediatric chest X-rays to detect the presence of tuberculosis. The computer aided diagnosis (CAD) tool consists of 4 phases: 1) lung field segmentation; 2) lung field subdivision; 3) feature extraction and 4) classification. Lung field segmentation is performed using a semi-automatic implementation of the active shape model algorithm. Two approaches to subdividing the lung fields into regions of interest are compared. The first divides each lung field into 21 overlapping regions of varying sizes, resulting in a total of 42 regions per image; this approach is called the big region approach. The second approach divides the lung fields into a large number of overlapping circular regions of interest. The circular regions have a radius of 32 pixels and are placed on an 8 x 8 pixel grid. This approach is called the circular region approach. Textural features are extracted from each of the regions using the moments of responses to a multiscale bank of Gaussian filters. Additional positional features are added to the circular regions."]},{"key":"dc:title","label":"Title","values":["Computer-aided diagnosis of tuberculosis in paediatric chest X-rays using local textural analysis"]}]}],"canonical_facts":{"dc:contributor.advisor":["Douglas, Tania S"],"dc:creator":["Mouton, Andre"],"dc:date.accessioned":["2014-07-28T18:17:35Z"],"dc:date.available":["2014-07-28T18:17:35Z"],"dc:date.issued":["2009"],"dc:description":["Includes abstract.","Includes bibliographical references (leaves 99-103)."],"dc:description.abstract":["This report presents a computerised tool to analyse the appearance of the lung fields in paediatric chest X-rays to detect the presence of tuberculosis. The computer aided diagnosis (CAD) tool consists of 4 phases: 1) lung field segmentation; 2) lung field subdivision; 3) feature extraction and 4) classification. Lung field segmentation is performed using a semi-automatic implementation of the active shape model algorithm. Two approaches to subdividing the lung fields into regions of interest are compared. The first divides each lung field into 21 overlapping regions of varying sizes, resulting in a total of 42 regions per image; this approach is called the big region approach. The second approach divides the lung fields into a large number of overlapping circular regions of interest. The circular regions have a radius of 32 pixels and are placed on an 8 x 8 pixel grid. This approach is called the circular region approach. Textural features are extracted from each of the regions using the moments of responses to a multiscale bank of Gaussian filters. Additional positional features are added to the circular regions."],"dc:identifier.uri":["http://hdl.handle.net/11427/3271"],"dc:language.iso":["eng"],"dc:publisher.department":["Department of Human Biology"],"dc:publisher.institution":["University of Cape Town"],"dc:title":["Computer-aided diagnosis of tuberculosis in paediatric chest X-rays using local textural analysis"],"dc:type":["Master Thesis"],"dc:type.qualificationlevel":["Masters"],"dc:type.qualificationname":["MSc"]},"updated_at":"2026-07-22T22:22:57Z"}