An Improved Technique for Computer Aided Detection of Breast Cancer using Mammograms and Image Mining
Abstract
Image mining focuses finding unusual patterns in images and deals with making association between different images from large image database. It deals with the extracting inherent and embedded knowledge, image data relationship, or other patterns and which is not explicitly found in the images. It is more than just an expansion of data mining to image domain. It is well known in the medical community that breast cancer is a leading cause of death in women (and to a lesser extent also affects men). When early detection of abnormalities is combined with proper medical attention, however, the risk of death and/or severe medical ramifications can be drastically reduced. Many devices and techniques (e.g., mammography) are currently under development to detect breast abnormalities earlier, and with greater accuracy than conventional devices. Presently Computer Aided Detection of breast cancer uses digital mammogram which has become the most effective technique for early breast cancer detection modality. Digital mammogram takes an electronic image of the breast and stores it directly in a computer. High quality mammogram images are high resolution and large size images. Processing these images require high computational capabilities. The transmission of these images over the net is sometimes critical especially if the diagnosis of remote radiologists is required. The aim of this study is to develop a research technique with Particle Swarm Optimization for feature subset selection and Association rule based classification to detect the mass whether normal, malignant or benign. It can be used to investigate the state of art of computer aided detection systems for digital mammograms, and evaluate the related techniques in image pre-processing, feature extraction and classification of digital mammograms. Furthermore, this paper will suggest the further research directions for next generation CAD for mammograms.
