We develop a model-based algorithm for the extraction of curvilinear structures in x-ray mammograms, corresponding to milk ducts, blood vessels, fibrous tissue, and spiculations. The diagnostic basis for the problem is sketched. Results of applying the algorithm are shown. The results accord closely with the CLS structures observed by Radiologists, and the algorithm has supported further processing to identify potential lesions and to match left and right breast x-ray pairs. A new perspective is provided on the debate on the relevant merits of film density images versus transmitted light.
CITATION STYLE
Cernea, N., & Brady, M. (1995). Finding curvilinear structures in mammograms. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 905, pp. 372–382). Springer Verlag. https://doi.org/10.1007/978-3-540-49197-2_47
Mendeley helps you to discover research relevant for your work.