CAD system provides an Medical assistance by scanning images from Computer Tomography for suspicious masses and highlights the segments like presence of tumors, neural blockage etc. Our project is used to improve the efficiency of CAD system and to detect the growth of tumor in efficient manner. This system is carried out by two processes. First, the image is preprocessed for the easier analysis and, second is the segmentation process, where the hybrid segmentation algorithm is used to detect the edge and create the database then the ROI are extracted from database. Finally the assessment of tumor growth is carried out using a nearest neighbor rule. This experimental result can help radiologist to improve the diagnosis efficiency by calculating the tumor growth in each stage accurately.