Interactive segmentation of abdominal organs from 3D CT and MRI images
Independent thesis Advanced level (degree of Master (One Year)), 20 credits / 30 HE creditsStudent thesis
Within the medical field, image segmentation is an important tool which can be used by radiologists and surgeons who want quantitative measurements of a lesion or organ. To be clinically useful, the tool has to be fast and easy to use. This work comprises implementation of the image foresting transform for segmentation using the Dijkstra algorithm and compares computation time between the implemented algorithm and a previous implemented algorithm, Bellman-Ford. These algorithms solve the shortest path with minimum cost problem. For a given cost function, similar results both in computation time and visual results were obtained with the two algorithms. Changing the cost functions, on the other hand, yielded very different segmentation results. The volume of liver and kidney was compared with manually delineated organs regarding seed planting and execution time. A graphical user interface has also been implemented.
Place, publisher, year, edition, pages
2010. , 42 p.
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-93510ISRN: LiU-ITN-TEK-A--10/025--SEOAI: oai:DiVA.org:liu-93510DiVA: diva2:628146
Subject / course