Show simple item record

dc.rights.licenseRestricted to current Rensselaer faculty, staff and students. Access inquiries may be directed to the Rensselaer Libraries.
dc.contributorYener, Bülent, 1959-
dc.contributorStewart, Charles V.
dc.contributorZaki, Mohammed J., 1971-
dc.contributor.authorBaskaran, Abhirami
dc.date.accessioned2021-11-03T08:07:06Z
dc.date.available2021-11-03T08:07:06Z
dc.date.created2014-04-14T11:15:03Z
dc.date.issued2013-12
dc.identifier.urihttps://hdl.handle.net/20.500.13015/1033
dc.descriptionDecember 2013
dc.descriptionSchool of Science
dc.description.abstractThe thesis builds a tool to detect cleft regions in the SMG based and characterize it in multiple domains. Implementing its detection, shape representation, feature extraction and classification provides a complete profiling and analysis of the cleft. Quantitative assessment of the clefts in both spatial and frequency domain adds more novelty to the study of branching morphogenesis. The shape changes of the clefts are tracked with the local scale features and a classifier is built to classify the clefts into their growth stages. Along similar lines the frequency domain features for the cleft regions were added to the shape feature based classifier by Bilgin et al. [3] to classify glands treated with ROCK inhibitors (inhibitors of branching). High classification accuracies of 97.5% and 94.61% were obtained proving the effectiveness of the local feature set. Global features at tissue scale were used to cluster growth factor induced SMGs. Growth factors affect the number of clefts formed and their deepening. The clustering results are in accordance to the biological perspectives and provide a clear insight into shape changes with growth factor treatment. The feature extraction and classification tool can be incorporated in models that simulate salivary gland morphogenesis to provide the necessary shape information for the simulations.
dc.description.abstractBranching morphogenesis is an important developmental mechanism in glands such as salivary glands, lungs, kidney and mammary glands [2]. Future breakthroughs such as building artificial glands to replace dysfunctional and diseased glands require a profound understanding of branching morphogenesis. Development of computer-aided techniques that can model branching patterns will provide a clear insight into the underlying mechanisms. Mouse submandibular salivary gland (SMG) provides good ground for the study of branching morphogenesis since it is easy to culture and exhibits well defined branching patterns. SMG exhibits boundary changes called clefts that begin as small indentations in the surface of the gland. These structures progress towards the interior of the gland finally forming branches.
dc.language.isoENG
dc.publisherRensselaer Polytechnic Institute, Troy, NY
dc.relation.ispartofRensselaer Theses and Dissertations Online Collection
dc.subjectComputer science
dc.titleMultiscale and multi-domain feature analysis of salivary gland morphogenesis
dc.typeElectronic thesis
dc.typeThesis
dc.digitool.pid170856
dc.digitool.pid170857
dc.digitool.pid170858
dc.rights.holderThis electronic version is a licensed copy owned by Rensselaer Polytechnic Institute, Troy, NY. Copyright of original work retained by author.
dc.description.degreeMS
dc.relation.departmentDept. of Computer Science


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record