Geometrijski deformabilni model za analizu medicinskih slika

scientific master's theses

• Author: Marko Subašić
• Mentor: Sven Lončarić (mentor)
• Committee member: Sven Lončarić (predsjednik povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2003, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering; Electronics
• Universal decimal classification (UDC): 621.3 - Electrical engineering

Abstract

Istraživacko podrucje rada je digitalna analiza medicinskih slika. Rad bavi se segmentacijom CT slika aneurizme abdominalne aorte. Motivacija za istraživanje je potreba za automatskom segmentacijom aneurizme abdominalne aorte, koja bi oslobodila lijecnike dugotrajne i zamorne rucne segmentacije. Segmentirati se može unutarnji i vanjski rub aorte. Pokazalo se da je segmentacija vanjskog ruba aorte puno veci problem od segmentacije unutarnjeg ruba, pa je glavno težište rada stavljeno na segmentaciju vanjskog ruba aorte. U prvom dijelu rada dan je kratak uvod u segmentaciju slika i deformabilne modele. Iznesena je osnovna teorija geometrijskih deformabilnih modela. U drugom dijelu rada opisana su cetiri pristupa segmentacije aneurizme abdominalne aorte, kojima je zajednicko korištenje geometrijskog deformabilnog modela. Svaki od pristupa na drugaciji nacin nastoji riješiti specificne probleme segmentacije vanjskog ruba aorte. Razvijene metode segmentacije isprobane su na nekoliko CT snimaka koje su se sastojale od nekoliko desetaka presjeka. Dva pristupa koja su pokazala najbolje rezultate testirana su na 11, odnosno 12 CT snimaka stvarnih pacijenata. Rezultati segmentiranja ta dva pristupa usporedeni su sa referentnim rezultatima koji su dobiveni rucnom segmentacijom.

Abstract (english)

The field of research of this thesis is in medical image analysis. The topic of the thesis is the segmentation of CT images of the abdominal aortic aneurysm. The research is motivated by the need for an automatic segmentation of the abdominal aortic aneurysm, that would relive physicians of tedious and time-consuming manual segmentation. Both the inner and the outer aortic border can be segmented. The segmentation of the outer aortic border presents a more difficult task than the inner aortic border segmentation. In first part of the thesis an introduction to image segmentation techniques and especially deformable model was presented. The theory of geometric deformable models is explained in more details. The second part of the thesis presents four approaches to the segmentation of abdominal aortic aneurysm. All approaches try to solve the problems of the outer aortic border segmentation in a different way, but they all use geometric deformable model as primary segmentation technique. Developed methods were tested on several CT scans. Two methods that gave best results were applied to 11 CT scans of real patients. The results were compared to reference segmentation results, obtained by manual segmentation.

• Keywords: digitalna obrada slike; digitalna analiza slike; digitalna obrada medicinskih slika; segmentacija; deformabilni modeli; level-set metoda; aneurizma abdominalne aorte
• Keywords (english): digital image processing; digital image analysis; medical image analysis; image segmentation; deformable models; level sets; abdominal aortic aneurysm
• Language: croatian
• URN:NBN: urn:nbn:hr:168:114312
• Study programme: Title: Postgraduate master programme in electrical engineering; Study programme type: university; Study level: postgraduate; Academic / professional title: Magistar znanosti elektrotehnike (Magistar znanosti elektrotehnike)
• Catalog URL: http://lib.fer.hr:8080/cgi-bin/koha/catalogue/detail.pl?biblionumber=29038
• Type of resource: Text
• Extent: 131 str. ; 30 cm
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-06-28 08:26:24