Radiologists generally uses mammogram images for extracting masses or cancer effected breast issues using texture features of the images by segmenting techniques. The most commonly used technique in this process is region of interest (ROI). But this method fails for large collection of mammogram image database. To address this issue this present paper proposed multi-ROI (M-ROI) technique. This method not only reduces limitations of ROI but also finds the very suitable texture features. This paper also evaluated the efficiency of the proposed M-ROI method using first order and second order statistical techniques.