Please correct the grammar of this report introduction

1i00Below is an introduction of an engineering coursework, could someone help me to correct my grammar?02i02br
02br
02br
02br
00Vessel extraction is one of the critical tasks in clinical practice. The physically complex vessel structure makes angiogram segmentation a challenging problem. The vasculatures are very important in planning and performing neurosurgery and cardiovascular surgery. They provides information on where the lesion draws, its blood supply and where its drains. They also serve as landmarks and guidelines to the lesion during surgery. Therefore, vessel extraction is a continuous research area in medical image analysis. A comprehensive review on vessel extraction techniques and algorithms can be found in [1].00 00In this paper, we focus on the new development of angiogram segmentation based on active contour model.02br
02br
02br
02br
00Deformable models, snakes, or active contours have been widely used in many applications, including edge detection, shape modeling, motion tracking, and image segmentation.00 00Snake model is first introduced by Kass in [2].00 00It evolves open or closed curves in an image, under the influence of the internal and external forces.00 00The internal forces are defined to ensure the integrity of the curve and to keep it from over bending, whereas the external forces are defined to lead the curve towards the target object.02br
02br
02br
02br
00Many advanced snake models are then introduced according to this initial model.00 00Most of them improve the overall performance of snake by introducing new formulation of the external forces, such as the gradient vector flow (GVF) snake in [3].02br
02br
02br
02br
00Classical snake models establish a very efficient framework for image segmentation or area-of-interest extraction by curve propagation. However, they cannot handle topological changes which are required in some applications.00 00In order to deal with these applications, the level set based deformable models are developed. Level set based model can handle complex object boundaries by flexible curve evolutions.00 00In addition, since the level set function is defined in the Euclidean space, it is easily extendable to higher dimensions.00 00The level set method is first introduced for front propagation by Osher and Sethian in [4]. 02br
02br
02br
02br
00This method is then applied in shape recovery and isolation of shape from its background by Malladi and Sethian in [5]. The concept provides a good alternative to the classical deformable models and has been widely used in medical image analysis.00 00There are number of level set based active contour models. They include the geometric active contour model [6] in terms of the mean curvature motion and the geodesic active contour model [7] which defines the problem in the Riemannian space. All of the above active contour models make use of the image gradient as their evolution and stopping criteria.00 00Since the image gradient is calculated by using the local curvature of the image intensity, this may probably lead the curve evolving towards the local minima. If the object is not precisely defined by its edges, the model may not extract the object correctly.02br
02br
02br
02br
00Chan and Vese [8] developed another type of active contour model using Mumford-Shah function [9] instead of using the image gradient. This model uses the global statistical measures of the object to adjust the level set function and fit the object with the zero level set. Consequently, the extracted object is not necessarily defined by its edges. This is useful in processing images which its target object boundaries are blurry, and the intensity of the object is assumed to be statistically homogenous.02br
02br
02br
02br
00In most clinical angiograms, however, the vessels are normally not homogenous in intensity. These may be due to several reasons including the limited exposure to minimize the side effects to the patient in X-ray images, the projection of the 3D object to 2D image through different depth, or the artifacts of the different imaging techniques.00 00Therefore, the existing active contour models using only global statistics are not adequate to extract the vessel boundary. In this paper, we propose a new approach for vessel extraction based on the Chan-Vese model [8] by incorporating a novel local term to enhance the segmentation results. This local term provides a simple yet robust statistical measure of the image contrast on the contrast enhanced image locally.02br
02br
02br
02br
00Two modes of operations of the proposed model are developed, namely the expansion mode and refinement mode.00 00The efficacy of the proposed model is demonstrated with both synthesized images and a number of clinical angiograms.00 00We compare the performance of the proposed model with Chan-Vese model, and a set of manually segmented images by clinicians.00 00The proposed model is found to provide better vessel extraction capability over the existing active contour models.02br
02br
02br
02br
00This paper is organized as follows:00 00In Section II, the concept and the formulation of the proposed model is described in details.00 00In Section III, the numerical implementation of the model is presented. In Section IV, several examples and demonstrations are shown. Finally, the conclusions are given in Section V.02br
02br
0-