Biomedical Engineering Reference
In-Depth Information
5
PARALLEL CO-VOLUME SUBJECTIVE
SURFACE METHOD FOR 3D
MEDICAL IMAGE SEGMENTATION
Karol Mikula
Department of Mathematics and Descriptive
Geometry, Slovak University of Technology,
Bratislava, Slovakia
Alessandro Sarti
Dipartmento di Elettronica, Informatica e
Sistemistica, University of Bologna, Bologna, Italy
In this chapter we present a parallel computational method for 3D image segmentation. It is
based on a three-dimensional semi-implicit complementary volume numerical scheme for
solving the Riemannian mean curvature flow of graphs called the subjective surface method.
The parallel method is introduced for massively parallel processor (MPP) architecture using
the message passing interface (MPI) standard, so it is suitable, e.g., for clusters of Linux
computers. The scheme is applied to segmentation of 3D echocardiographic images.
1.
INTRODUCTION
The aim of segmentation is to find boundaries of an object in an image. In a
generic situation these boundaries correspond to edges. In the presence of noise,
which is intrinsically linked to modern noninvasive acquisition techniques (such
as ultrasound), the object boundaries (image edges) can be very irregular or even
interrupted. The same happens in images with occlusions, subjective contours (in
