Medical Imaging - Boundary Solutions

Abstract

Boundary detection to a narrow scale or for a Region of Interest are required for studying the affected area in an human organ depending on the nature of the disease and the damage it caused to the specific organ. The problem is narrowed down to edge detection and the associated complexities. Precisely, it concentrates over a small region of interest confined to a specific area lying anywhere on the shape of study under consideration. Even though there are many imaging methods which help to overcome these kinds of situation, there are limitations. This paper addresses these issues with the help of Contourlet Transformation. Further, Gradient and Laplacian operators help in tuning the edge detection. Comparatively, the proposed methods perform better than the traditional methods. But, still the direction specific issues and extension issues made these techniques difficult to achieve the expected accuracy. Moreover, the Contourlet transform addresses the edge detection problem very well in digital domain. Finally, the Contourlet Transformation helped to overcome all of these issues by capturing the required data that involved the features in an image which ultimately focused on bringing the discreteness of the nature of the problem.