Our aim is to compare two mammograms (left-right, temporal) in an unsupervised manner. To this end, we propose a novel region matching algorithm (RMA) for mammograms based upon the non-emergence and non-enhancement of maxima and the causality principle of integral invariant scale space (in a limited sense). The algorithm has several advantages over commonly used methods for comparing segmented regions as shapes. First, it gives improved key-points alignment for optimal shape correspondence. Second, it identifies new growths and complete/partial occlusion in corresponding regions by dividing the segmented region into sub-regions based upon the extrema that persist over all scales. Third, the algorithm does not depend upon the spatial locations of mammographic features and eliminates the need for registration to identify salient changes over time. Finally, the algorithm is fast to compute and requires no human intervention. © 2012 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Janan, F., & Brady, S. M. (2012). Region matching in the temporal study of mammograms using integral invariant scale-space. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7361 LNCS, pp. 173–180). https://doi.org/10.1007/978-3-642-31271-7_23
Mendeley helps you to discover research relevant for your work.