Abstract
This paper develops an approach aiming to identify a pattern defined by malignant dermoscopic skin lesions presented as lesion feature vectors (LFVs) in a 4D Riemannian manifold. The manifold consists of all 4-tuples and its metric is defined on the basis of the Total Dermoscopy Score (TDS) formula used in the ABCD diagnosis rule. Tools in Riemannian geometry including distance functions, directional angles, polar, cylindrical and spherical coordinates are used to study the geometric structures of a sample space defined by malignant 4D LFVs. To explore the geometric structures and the distribution pattern of the malignant LFVs, we find methods to visualize the objects in a 4D manifold by projecting them onto a 2D or 3D space via polar, generalized cylindrical and spherical coordinates. To observe malignancy identification structures in the newly constructed manifold a data-set of 70 lesion images with a ground truth were used to generate LFVs. To build a surface separating the benign and malignant LFVs a linear TDS-based and a non-linear support vector machine (SVM) classifiers were applied. The SVM is build with a polynomial kernel, whose degree and parameters were suggested by a geometric structure observed in a 3D space. A statistical comparison, on the base of experimental results, showed that the polynomial SVM has a better f-measure accuracy than the linear TDS based one.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
American Cancer Society. Cancer Facts & Figures (2010), http://www.cancer.org/ (accessed July 26, 2010)
Argenziano, G., Fabbrocini, G., Carli, P., De Giorgi, V., Sammarco, E., Delfino, M.: Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions. Comparison of ABCD rule of dermatoscopy and a new 7-point checklist based on pattern analysis. Arch. Dermatol. 134, 1563–1570 (1998)
Argenziano, G., Soyer, H.P., De Giorgi, V., et al.: Interactive atlas of dermoscopy. EDRA Medical Pub., Milan (2000)
Mete, M., Sirakov, N.M.: Application of active contour and density based models for lesion detection in dermoscopy images. BMC Bioinformatics 11(suppl. 6), S23 (2010), doi:10.1186/1471-2105-11-S6-S23
Mete, M., Sirakov, N.M.: Dermoscopic diagnosis of melanoma in a 4D feature space constructed by active contour extracted features. Journal of Medical Imaging and Graphics 36, 572–579 (2012)
Mulchrone, K., Choudhury, K.: Fitting an ellipse to an arbitrary shape: implications for strain analysis. J. of Structural Geology 26(1), 143–153 (2004)
O’Neill, B.: Semi-Riemannian geometry with applications to relativity. Academic Press, New York (1983)
Nachbar, F., Stolz, W., Merkle, T., et al.: The ABCD rule of dermatoscopy: High prospective value in the diagnosis of doubtful melanocytic skin lesions. J. of the American Academy of Dermatology 30(4), 551–559 (1994)
Petersen, P.: Riemannian geometry. Graduate Texts in Mathematics, vol. 171. Springer, New York (1998)
Scholkopf, B., Smola, A.J., Williamson, R.C., Bartlett, P.L.: New Support Vector Algorithms. Neural Comput. 12, 1207–1245 (2000)
Sirakov, N.M., Mete, M., Nara, S.C.: Automatic boundary detection and symmetry calculation in dermoscopy images of skin lesions. In: IEEE ICIP 2011, Brussels, pp. 1637–1640 (2011)
Sirakov, N.M., Ushkala, K.: An Integral Active Contour Model for Convex Hull and Boundary Extraction. In: Bebis, G., Boyle, R., Parvin, B., Koracin, D., Kuno, Y., Wang, J., Pajarola, R., Lindstrom, P., Hinkenjann, A., Encarnação, M.L., Silva, C.T., Coming, D. (eds.) ISVC 2009, Part II. LNCS, vol. 5876, pp. 1031–1040. Springer, Heidelberg (2009)
Thieu, Q.T., Luong, M., Rocchisani, J.-M., Viennet, E.: A Convex Active Contour Region-Based Model for Image Segmentation. In: Real, P., Diaz-Pernil, D., Molina-Abril, H., Berciano, A., Kropatsch, W. (eds.) CAIP 2011, Part I. LNCS, vol. 6854, pp. 135–143. Springer, Heidelberg (2011)
Vapnik, V.: The Nature of Statistical Learning Theory. Springer, New York (1995)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Mete, M., Ou, YL., Sirakov, N.M. (2012). Skin Lesion Feature Vector Space with a Metric to Model Geometric Structures of Malignancy for Classification. In: Barneva, R.P., Brimkov, V.E., Aggarwal, J.K. (eds) Combinatorial Image Analaysis. IWCIA 2012. Lecture Notes in Computer Science, vol 7655. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-34732-0_22
Download citation
DOI: https://doi.org/10.1007/978-3-642-34732-0_22
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-34731-3
Online ISBN: 978-3-642-34732-0
eBook Packages: Computer ScienceComputer Science (R0)