THE DEVELOPMENT OF CLUSTERING AND CLASSIFICATION TECHNIQUES FOR MAMMOGRAM IMAGES
|
Breast cancer is among the leading causes of cancer related deaths among women. Early detection represents a very important factor in cancer treatment and allows reaching a high survival rate. Mammograms are considered the most reliable method in early detection of cancer. Digital mammograms are among the most difficult medical images to be read due to their low contrast and differences in the types of tissues. Computer aided diagnosis systems may prove useful for doctors for interpreting mammograms. This improves the breast cancer prediction accuracy. Image processing techniques and data mining techniques play important roles in creating a successful computer aided diagnosis system. Image processing techniques are used as a preprocessing step because of low contrast of images, hard to read masses in images. Data mining techniques (classification and clustering techniques) need features, hence, features are created from these images act as an input for these data mining techniques. These data mining techniques analyze these images by using their features. In this project, we will study various image processing techniques and data mining techniques for analyzing digital mammograms. We will develop a computer aided diagnosis system that may help radiologists in interpreting mammograms.
|