February 22, 2008
Speaker: Jung-Ha An
Title: Generating Prior Shapes and Simultaneous Image Segmentation and
Registration with an Application to Medical Imaging
Abstract: Generating prior shapes and image segmentation and
registration are important but challenging research in medical
imaging. In medical applications, prior shape plays a key
role in obtaining accurate result in image segmentation. A new
algorithm for generating prior shape models using a Self-Organizing map
is presented. The aim of the model is to develop an approach for
shape representation and classification to detect differences in the
shape of anatomical structures. The presented algorithm is
applied to the eighty-five two-chamber normal human heart ultrasound
end-diastolic cardiac borders. The experimental results show the
effectiveness of the algorithm in generating shape representation and
classification of given various human heart cardiac borders. For
a simultaneous image segmentation and non-rigid registration, a new
variational partial differential equation edge based level set method
using prior shape and intensity information is presented. The
segmentation is obtained by finding a non-rigid registration to the
prior shape. The non-rigid registration consists of both a global rigid
transformation and a local non-rigid deformation. In this model,
a prior shape is used as an initial contour which leads to decrease the
numerical calculation time. The models is tested against two
chamber end systolic ultrasound images from thirteen human
patients. The experimental results provide preliminary evidence
of the effectiveness of the model in detecting the boundaries of the
incompletely resolved objects which were plagued by noise, dropout, and
artifact.