Segmentation of mammograms in topographic approach

This paper presents a novel approach to finding a priori interesting regions in mammograms. We propose a topographic representation called the iso-level contour map consisting of iso-level contours at multiple intensity levels, in order to delineate those regions of interest (ROIs) in mammograms which appear to be salient. Such salient areas form a dense quasi-concentric pattern of iso-level contours and these are made explicit in the topographic representation. The algorithm detects object boundaries by analysing geometrical and topological properties of iso-level contours. The "saliency" of a region is measured topologically as the minimum nesting depth of contours using an inclusion tree that is a hierarchical representation of the enclosure relationship between contours. The result is an analysis of the global structure of objects in mammograms. Experimental results are presented.