Semi-automated region of interest selection tool for mammographic image

Segmenting out region of interest (ROI) accurately for gray level and texture analysis in mammography has always been a challenge because features within mammographic image do not normally have a well defined boundary. Manual hand drawn selection method is too cumbersome while using a fully automated method means we will lose flexibility and control over the ROI selection process. This means the ROI selection may not be as accurate as we would like it to be. Therefore, this paper proposes a semi-automated tool in ROI selection which is more efficient than hand drawn method while still maintaining a significant amount of human control compared to a fully automated method. This tool has been tested on a number of samples and we found that it is intuitive and simple to use but still able to give accurate ROI selection.