3D Prostate Surface Detection from Ultrasound Images Based on Level Set Method

Abstract

Abstract. Accurate detection of prostate boundaries is required in many diagnostic and treatment procedures for prostate diseases. In this paper, a new approach based on level set method to perform 3D prostate surface detection from transrectal ultrasound (TRUS) images is presented. Contrary to many other deformable models, level set method offers several advantages such as minimal need for user input, flexible topology, and straightforward extension to 3D. However, it is subject to "boundary leaking" problem for ultrasound image segmentation due to the poor image quality. In this work, we first develop a fast discrimination method to extract the prostate region, then this region information, instead of the spatial image gradient, is incorporated into the level set method to remedy the "boundary leaking" problem. Various experimental results show the effectiveness of the proposed method.

Segmentation of prostate from ultrasound images using level sets on active band and intensity variation across edges

“…Patch size Edge descriptor matching patches are the patches having the best displacement defined in Eq. (14). The normalized cross-correlation has been commonly used as a metric to evaluate the similarity (dissimilarity) between two images.…”

Prostate segmentation in ultrasound images with deformable shape priors

“…Even though more dense acquisition and 3D volume acquisition have been suggested [11] and some 3D prostate segmentation algorithms are also being developed [8,10,12,13], most clinics still use sparse data for routine clinical tasks. The proposed segmentation method consists of two steps: offline model training and online prostate segmentation with prior model constraints.…”

Level set segmentation of bovine corpora lutea in ex situ ovarian ultrasound images

“…Our study represents the first investigation of semi-automated segmentation of corpora lutea from ultrasound images. The level set method lends itself to image segmentation tasks because it requires minimal user input, accommodates arbitrary changes in region topology and offers a straightforward extension to higher dimensional data [ 3 ]. Figure 1(a) shows an image of a CL and Figure 1(b) shows the desired segmentation result as drawn by a human expert in ovarian ultrasound interpretation.…”

“…Level set methods has been proven effective both in general [ 17 - 20 ] and for ultrasound image segmentation [ 21 , 22 ]. Recent applications of the level set method to prostate ultrasound image segmentation [ 3 , 7 , 23 ] show that this technique is accurate and flexible (it can handle contours of varying shape, size, and concavity). A level set method was chosen for the current study because of the prior success of level set methods in prostate segmentation and its ability to easily handle arbitrary changes in contour topology.…”

“…Fan et al [ 3 ] used a level set method to perform three-dimensional (3D) surface detection from ultrasound images of the prostate. A fast discrimination method was used to roughly extract the prostate region.…”