New tumor detection method using Nl-means filter and histogram study

In this article we treat a very interesting and very important research subject that affects directly the human brain. The subject here is the tumor detection on a brain MRI The tumors evaluation, its progressions and its area, are very important data to assist the doctors in the disease diagnosis. However, tumor observation and image analysis proves to be a delicate exercise requiring a lot of MRI and using multiple acquisitions under different protocols. The hard challenge now is to accurately detect the tumor surface. In this article we present a new method of tumor regions detection on brain MRI. This method consists of three steps: a) filtering using Nl-means filter and extracting the brain from skull using EMBE method b) the MRI image study and histogram analysis, this step will help us to find a threshold of intensity to extract tumor; c) the tumor extraction. We will complete our work by a tumor characterization and determination of the geometric properties.