This paper describes approaches and methods that are aimed to solve various medical image processing tasks, including extraction of region of interest, the extracted regions contour processing and the following 3D visualization of the extracted region. Two different approaches for control point acquisition and 3D model reconstruction using the points acquired from the region of interest are proposed. The proposed methods of 3D visualization use a vector-based surface model creation rather than raster-based systems to significantly improve computational efficiency and the quality of generated 3D models. The proposed methods were tested on the medical images of a brain acquired by computer tomography and proven to be applicable to different types of region of interest. The results can be used to provide practical improvements to the reliability of medical diagnostics.