| Detection of small brain tumors in MRI images may cause more diagnostic difficulty than larger tumors, and therefore, increases the possibility of human error when analyzed by a specialist. Automation of the detection process not only minimizes human error but also speeds up the time consuming process of evaluating large number of MRI images. In this work, we propose a new hybrid methodology that utilizes image processing techniques, content based analysis, and statistical analysis to detect, locate and mark brain tumors. Our main contribution is design and development of an efficient and accurate algorithm capable of detecting small and subtle tumors. |
