An important step for computerized analysis of breast magnetic resonance imaging (MRI) is segmentation of the breast region. Due to the similar signal intensity of broglandular tissue and the chest wall, the segmentation process is difficult for breasts with broglandular tissue connected to the chest wall. In order to overcome this challenge, a new framework is presented that relies on a chest region atlas. The proposed method rst detects the approximated breast{chest wall boundary using an intensity-based operation. A support vector machine (SVM) then determines the connectivity of broglandular tissue to the chest wall by the extracted features from the obtained breast{chest wall boundary. Finally, the obtained breast{chest wall boundary is accurately re ned using the geometric shape of the chest region, which is obtained by an atlas-based segmentation method. The proposed method is validated using a dataset of 5964 breast MRI images from 126 women. The Dice similarity coefficient (DSC), total overlap (TO), false negative (FN), and false positive (FP) values are calculated to measure the similarity between automatic and manual segmentation results. Our method achieves DSC, TO, FN, and FP values of 96.46%, 96.41%, 3.59%, and 3.51%, respectively. The results prove the effectiveness of the presented algorithm for breasts with different sizes, shapes, and density patterns.