Segmentation and quantitative characterization of breast masses imaged using digital mammography
| dc.contributor.advisor | Rae, William Ian Duncombe | en_ZA |
| dc.contributor.author | Nkwenti, Sussan Acho | en_ZA |
| dc.date.accessioned | 2019-07-22T09:12:13Z | |
| dc.date.available | 2019-07-22T09:12:13Z | |
| dc.date.issued | 2018 | en_ZA |
| dc.description | Thesis (Ph.D.(Medical Physics))--University of the Free State, 2018 | en_ZA |
| dc.description.abstract | 𝑬𝒏𝒈𝒍𝒊𝒔𝒉 Breast cancer is the leading cause of cancer death among women. Screening Mammography is the most effective method currently available for early detection of breast cancer. When breast cancer is detected at an early stage the prognosis is good because the tumour is smaller and more often well-differentiated, and less likely to have spread to regional lymph nodes. Computed radiography and direct digital detector mammography imaging systems provide a wide dynamic range for proper display of different densities of breast tissue areas. Their response over a wide range of X-ray intensities is linear; consequently, small differences between the attenuation coefficients of breast structures over a wide range of densities are clearly displayed. This includes the low signal areas associated with high densities found within tumours. Some masses infiltrate the surrounding breast tissue hence they exhibit ill- defined and intensity inhomogeneous boundaries with rough contour, while other masses exhibit well-defined edges and in most cases they possess smooth, round or oval shapes with macro-lobulations. The morphologic features of a mass such as its shape, margin and density give a clue to its benign or malignant nature. This study investigates and quantifies the changes in shape-based descriptors due to changes in the location of the initial level set contour in region based active contour models in delineating mammographic masses and proposes new methods to eliminate contour leakage and contour traps in active contour segmentation models which are due to intensity inhomogeneity within tumours and boundary regions of tumours. Furthermore, the study proposes a contextual region of interest model to assess the variation of texture features from the core to its periphery of biopsy proven malignant masses as a concept of tumour modelling in mammography and also the variation of texture features between grade 2 and grade 3 masses as a concept of tumour grading in mammography with texture analysis. ___________________________________________________________________ | en_ZA |
| dc.description.abstract | 𝑨𝒇𝒓𝒊𝒌𝒂𝒂𝒏𝒔 Borskanker is die hoofoorsaak van kankersterftes onder vroue. Sifting-mammografie is die mees doeltreffende metode wat tans beskikbaar is vir vroeë opsporing van borskanker. Wanneer borskanker in 'n vroeë stadium opgespoor word, is die prognose goed omdat die gewas kleiner en meer dikwels goed gedifferensieerd is, en minder geneig is om na plaaslike limfknope te versprei. Rekenaarradiografie en direkte digitale detektor mammografie-beeldingstelsels bied 'n wye, dinamiese reeks vir behoorlike vertoon van verskillende digthede van borsweefselareas. Hul reaksie oor 'n wye reeks X-straalintensiteite is lineêr; gevolglik word klein verskille tussen die verswakkingskoëffisiënte van borsstrukture oor 'n wye reeks digthede duidelik vertoon. Dit sluit die lae seinareas in wat verband hou met hoë digthede wat binne gewasse gevind word. Sommige massas infiltreer die omliggende borsweefsel, daarom vertoon hulle swak gedefinieerde en intensiteit inhomogene grense met growwe kontoer, terwyl ander massas goed gedefinieerde rante vertoon en in die meeste gevalle het hulle gladde, ronde of ovaalvorms met makro-lobulasies. Die morfologiese kenmerke van 'n massa soos die vorm, marge en digtheid gee 'n leidraad tot die goedaardige of kwaadaardige aard daarvan. Hierdie studie ondersoek en kwantifiseer die veranderinge in vorm-gebaseerde beskrywers as gevolg van veranderinge in die ligging van die aanvanklike vlak vasgestelde kontoer in streekgebaseerde aktiewe kontoermodelle in die afbakening van mammografiese massas en stel nuwe metodes voor om kontoerlekkasie en kontoervalle in aktiewe kontoersegmenteringsmodelle uit te skakel wat te wyte is aan intensiteit inhomogeniteit binne gewasse en grensstreke van gewasse. Verder stel die studie 'n kontekstuele belangstellingsmodel voor om die variasie van tekstuurkenmerke van die kern tot die periferie van biopsie-bewese kwaadaardige massas te bepaal as 'n konsep van tumormodellering in mammografie en ook die variasie van tekstuurkenmerke tussen graad 2- en graad 3-massas as 'n konsep van tumorgradering in mammografie met tekstuuranalise. ___________________________________________________________________ | af_ZA |
| dc.description.abstract | 𝑺𝒆𝑺𝒐𝒕𝒉𝒐 Mofetshe wa matswele ke le leng la mafu a itlhommeng pele a bolayang basadi. Ho hlahloba ka seipone se shebang mofetshe wa matswele ke mokgwa o thusang haholo nakong ena mme o ka bontsha mofetshe wa matswele ka nako. Ha mofetshe wa matswele o bonahetse ka nako, maikutlo a dingaka a ka ba matle hobane hlahala e sa le nyenyane mme hangata e sa ntse e ka lemohwa hantle. Menyetla ya ho ikatisa ha e ya sebakeng sa methapo, e sa ntse e fokola. Seipone se sebetsang ka khomphutha ke sehlahlobisi sa maranrang se hlahlobang mofetshe wa matswele le tsa boinahanelo tse ka thusang ka tse fapaneng tse pharaletseng tse bontshang tse fapaneng mabapi le ho fokotseha ha dihlahala tsa matswele. Ho arabela seipone se shebang ka botebo le bophara ho sebetsa ka kotloloho. Kahoo diphapano tse fokolang tsa ho se bonahale hantle tsa mofetshe wa matswele hodima sebopeho se pharaletseng di bonahala habobebe ka mokgwa o hlakileng. Tsena di kenyeletsa dikarolo tse bontshang hanyenyane le tse bontshang haholo ka hara dihlahala. Tse ding tse ngata tse kenelletseng matsweleng, di bontsha ntho tse sa hlokomeleheng habobebe le matla a meedi e se nang sebopeho se tshwanang se nang le maqhutsu. Ha e le tse ding tse ngata di bontsha dikarolo tse bonahalang tsa ho qetela ka dinako tse ding di hlaha di bataletse, di le sekolokoto kapa di le motopo ho ya ka matshwao a mofetshe wa matswele. Dibopeho tse qalang tsa mofetshe wa matswele jwalo ka sebopeho sa moo a fellang a fana ka setshwantsho sa ho qala ha yona le ho tswa taolong. Boithuto bona bo sibolotse le ho hlophisa ho ya ka ditlhaloso tsa dibopeho tse fetohang tsa se hlaloswang, tse fetolwang ke ho fetoha ha sebaka sa moo e neng e fumaneha qalong mme e bonahala haholo mokgwa wa sebopeho ho hlalohanya ka seipone sa ho hlahloba mofetshe wa matswele le ho hlahisa mekgwa e metjha ya ho fedisa ho dutla le ho itlhahisa ka mokgwa wa dikarolwana tse bakwang ke boima ke ho se tshwane ka hara dihlahala le dibakeng tseo dihlahala di fumanehang ho tsona. Ho feta moo, boithuto bona bo sisinya hore ho sebetswe ho ikamahantswe le lewa la sebaka ho hlahloba dihlahala tse sa tshwaneng ho tloha qalong ho isa ho tse seng di itjadile, diqhalane mme diqalella ho etsa hore sesole sa mmele se jane. Ena e ba taba ya dihlahala tse bontshwang ka seipone sa mofetshe wa matswele le ho fapana ha matshwao a dinama pakeng tsa tlhopho ya bobedi le tlhopho ya boraro ya mofetshe wa matswele. Ena ke taba ya ho hlophisa dihlahala seiponeng sa mofetshe wa matswele e le ho lemoha sebopeho sa dihlahala tseo. ___________________________________________________________________ | st_ZA |
| dc.description.abstract | 𝑰𝒔𝒊𝒁𝒖𝒍𝒖 Umdlavuza webele uyimbangela ehamba phambili yokufa kwabantu besifazane bebulawa umdlavuza. Ukuhlola kweMammography iyona ndlela etholakalayo esebenza kahle njengamanje yokushesha kubanjwe umdlavuza webele kusenesikhathi. Uma umdlavuza webele utholwa kusenesikhathi ukubhebhetheka akukho ngoba isimila sincane futhi sivame ukuhluka kahle, futhi mancane amathuba okuthi sisabalale kumaseli amancane amise okwabhontshisi. Iradiography yekhompyutha kanye nezinhlelo zedijithali zokuthwebula zemammography zinikeza ububanzi obuguquguqukayo bokuboniswa okufanele kokuminyana okuhlukahlukene kwezindawo zezicubu zamabele. Ukuphendula kwawo kwizinhlobonhlobo eziningi zama-inthensithi eX-ray kuqondile; ngenxa yalokho, umehluko omncane phakathi kwezinombolo okuphindaphindwa ngazo wokunciphisa wezakhiwo zamabele phezu kobubanzi obuhlukahlukene bokuminyana uboniswa ngokucacile. Lokhu kufaka phakathi izindawo zesignali ephansi ezihlobene nokuminyana okuphezulu okutholakala ngaphakathi kwezimila. Okunye kungena ezicutshini zebele ezizungezile yingakho kukhombisa ukugula kanye nokuqina okungafani nendawo emagebhugebhu, kuyilapho ezinye izindawo zibonisa imiphetho ephila kahle engenakho ukugula futhi ezimweni eziningi zinezimo ezibushelelezi, eziyindilinga noma eziyisiyingi ezinamamacro-lobulations. Izici zemofoloji zesisindo esifana nokuma kwayo, umkhawulo kanye nokuminyana zinikeza inkomba yemvelo yayo eyingozi. Lolu cwaningo luphenya futhi lulinganise izinguquko ezincazelweni ezisuselwe ekumeni ngenxa yezinguquko endaweni yezinga lokuqala lekhonta esethiwe esifundeni esisekelwe kumamodeli asebenzayo wekhonta ekudwebeni ubuningi bemammographic futhi luphakamisa izindlela ezintsha zokuqeda ukuvuza kwekhonta kanye nezicupho zamakhonta kumamodeli asebenzayo wokuhlukaniswa kwekhonta okubangelwa ukushuba kokungafani ngaphakathi kwezimila kanye nezindawo eziwumngcele wamathumba. Ngaphezu kwalokho, ucwaningo luhlongoza isifunda somongo semodeli yentshisekelo yokuhlola ukuhlukahluka kwezici zokuthungwa kusukela engqikithini kuya endaweni yayo yobuningi obuyingozi obufakazelwe ukususwa kwamaseli njengohlelo lokumodela isimila kwimammography kanye nokuhlukahluka kwezici zekhwalithi phakathi kwebanga lesi-2 nebanga lesi-3 njengohlelo lokuhlelwa kwesimila kumammography ngokuhlaziywa kwekhwalithi. ___________________________________________________________________ | isiZ_ZA |
| dc.identifier.uri | http://hdl.handle.net/11660/10152 | |
| dc.language.iso | en | en_ZA |
| dc.publisher | University of the Free State | en_ZA |
| dc.publisher | Abstract in other languages 𝘚𝘤𝘳𝘰𝘭𝘭 𝘥𝘰𝘸𝘯 𝘧𝘰𝘳 𝘈𝘧𝘳𝘪𝘬𝘢𝘢𝘯𝘴, 𝘚𝘦𝘚𝘰𝘵𝘩𝘰 𝘢𝘯𝘥 𝘐𝘴𝘪𝘡𝘶𝘭𝘶 | |
| dc.rights.holder | University of the Free State | en_ZA |
| dc.subject | Breast cancer | en_ZA |
| dc.subject | Mammography | en_ZA |
| dc.subject | Early detection | en_ZA |
| dc.subject | Computed radiography | en_ZA |
| dc.subject | Direct digital detector mammography imaging systems | en_ZA |
| dc.title | Segmentation and quantitative characterization of breast masses imaged using digital mammography | en_ZA |
| dc.type | Thesis | en_ZA |