Performance analysis on three dimensional surface reconstruction of head magnetic resonance images

In MRI images, the boundary of an encephalic tissue is highly curved and irregular. Three dimensional reconstruction of such encephalic tissue is complicated. The surface reconstruction is the sub-field of Medical imaging which provides an effective way to investigate and determine brain related diseases in an efficient and effective manner. The basic purpose of 3-D surface reconstruction is to analyze the brain images precisely in order to effectively diagnose and examine the diseases for surgical planning and tumor localization. Reconstruction of tumor images is the goal in dealing with these images. In this paper, a brief overview is given on the advantages and disadvantages of existing surface reconstruction methods in clinical applications. The traditional cube based algorithms extracts the surface by forming imaginary cube and then determines the polygons needed to represent the part of the isosurface that passes through this cube. But, it requires post processing and needs more Computational time for reconstruction. Also they cannot provide the proof of correctness. The vector machine based algorithms like Immune Sphere Shaped Support Vector Machine (ISSSVM) transforms the highly irregular object into the high dimensional feature space and construct the hyper-sphere as compact as possible which encloses almost all the target object. This paper concludes that ISSSVM can outperform the cube based algorithm by reconstructing the irregular boundaries of the encephalic tissue efficiently without post processing. It can also provide its proof of correctness with greater accuracy.