Abstarct: This years, early detection and diagnosis of breast cancer is one of the most important issues for radiologists. Early detection of breast cancer can increase the life expectancy of patients. Using mammograms is currently the most popular method for detection of breast cancer. Micro Calcifications or MCs are tiny deposits of calcium that appear as clusters of bright spots in mammograms. Clustered MCs can be an important indicator of breast cancer. They appear in 30%–50% of cases diagnosed by mammographic screenings. Computer aided detection or diagnosis (CAD) systems, which use computer technologies to detect abnormalities in mammograms such as calcifications, masses, and architectural distortion, and the use of these results by radiologists for diagnosis can play a key role in the early detection of breast cancer and help to reduce the death rate among women with breast cancer. In this research we analyzed and compared six methods for detecting the position of MCs in mammograms and proposed a intensity-region baxsed method for segmentation of detected MCs in a mammogram. The methods for MC detection differ in the two main steps of designing a CAD system. These two steps are feature extraction and classification. We used three different feature vectors for MC detection. We design CAD systems using the proposed feature vectors and compare the power of these methods. The first feature vector is consisted of the raw intensity of pixels within a window centered at the location of interest. The second and third feature vectors are consist of frequency and statistical features. We used the two classifiers SVM (support vector machine) and NN (neural network) for classifying the pixels in a mammogram using each of three feature vectors. Accordingly six CAD systems using and three feature vectors and two classifiers are constructed. Our experiments on these CAD systems shows that using frequency and statistical features with SVM classifier provides the best recognition rate among these CAD systems.