Abstract
In medical field visualization of the organs is very imperative for accurate diagnosis and treatment of any disease. Brain tumor diagnosis and surgery also required impressive 3D visualization of the brain to the radiologist. Detection and 3D reconstruction of brain tumors from MRI is a computationally time consuming and error-prone task. Proposed system detects and presents a 3D visualization model of the brain and tumor inside which greatly helps the radiologist to effectively diagnose and analyze the brain tumor. We proposed a multi-phase segmentation and visualization technique which overcomes the many problems of 3D volume segmentation methods like lake of fine details. In this system segmentation is done in three different phases which reduces the error chances. The system finds contours for skull, brain and tumor. These contours are stacked over and two novel methods are used to find the 3D visualization models. The results of these techniques, particularly of interpolation based, are impressive. Proposed system is tested against publically available data set [41] and MRI datasets available from MRI & CT center Rawalpindi, Pakistan [42].