Please use this identifier to cite or link to this item:
|Title:||Level set method for biomedical images||Authors:||Annamalai Lakshmanan||Keywords:||DRNTU::Engineering::Electrical and electronic engineering||Issue Date:||2006||Abstract:||The introduction of advanced imaging technologies has increased the number of images that a clinician or biologist has to handle. These imaging technologies are used for a variety of applications such as image guided surgery, image - based diagnosis, image visualization etc., which necessitates a need for automatic or semi-automatic technique for the analysis of medical images. Even though medical imaging plays a vital role in various applications, the community of medical image analysis has been surrounded by the challenging problem of extracting useful information about the anatomic structures from the large amount of data due to the tremendous variability of object shapes and the variation in image quality due to noise and sampling artifacts. With the amount of medical data increasing, manual segmentation will not be feasible thus necessitating a robust segmentation algorithm to obtain accurate, repeatable and quantitative results. In this work we focus on the level set method for finding the object of interests in medical images, in particular, the confocal microscopic images and the magnetic resonance (MR) images. The work discusses in detail the foundation of level set method and the various researches that have been carried out in the recent years. The work also talks about the choice of this method for the application in hand by comparing the results obtained for various methods with the results obtained using level set method. Finally the work discusses about the incorporation of active contour without edges in the fast level set without solving partial differential equation (PDE) framework, which is done in order to reduce the computational burden in the traditional level set method and to detect objects that are very close to each other in addition, it uses the advantages of active contour without the edges to detect the objects that are not defined by gradients.||Description:||74 p.||URI:||http://hdl.handle.net/10356/39118||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||EEE Theses|
Page view(s) 5249
checked on Oct 21, 2020
checked on Oct 21, 2020
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.