Image processing plays a vital role in the detection of medical images segmentation. Accurate
detection will help the radiologilist to predict the images masses. Using Deep Learning
methods the segmentation of medical images can be possible with best results when compared
with state o art methods. In this paper we have compared the deep learning methods like a) U
net, b) U-net attention, c) Dense-net, d) Attention Dense U-net for the segmentation of
mammography images. Performance metrics like Accuracy, Sensitivity, Specificity, F1 Score,
and AUC are measure for the above mentioned models. Attention Dense U-net outperforms
the best in accuracy with 78.38%, sensitivity is 77.89%, F1 score with 82.24 , and AUC with
86.05%.