Segmentation of Clinical Endoscopic Images Based on the Classification of Topological Vector Features
https://doi.org/10.18255/1818-1015-2013-6-162-173
Abstract
In this work, we describe a prototype of an automatic segmentation system and annotation of endoscopy images. The used algorithm is based on the classification of vectors of the topological features of the original image. We use the image processing scheme which includes image preprocessing, calculation of vector descriptors defined for every point of the source image and the subsequent classification of descriptors. Image preprocessing includes finding and selecting artifacts and equalizating the image brightness. In this work, we give the detailed algorithm of the construction of topological descriptors and the classifier creating procedure based on mutual sharing the AdaBoost scheme and a naive Bayes classifier. In the final section, we show the results of the classification of real endoscopic images.
About the Authors
O. A. Dunaeva
P. G. Demidov Yaroslavl State University
Russian Federation
Russian Federation
канд. физ.-мат. наук, доцент каф. вычислительных и программных систем ЯрГУ, Международная лаборатория «Дискретная и вычислительная геометрия» им. Б. Н. Делоне,
Sovetskaya str., 14, Yaroslavl, 150000, Russia
D. B. Malkova
P. G. Demidov Yaroslavl State University
Russian Federation
Russian Federation
аспирант, Международная лаборатория «Дискретная и вычислительная геометрия» им. Б. Н. Делоне,
Sovetskaya str., 14, Yaroslavl, 150000, Russia
M. L. Machin
P. G. Demidov Yaroslavl State University
Russian Federation
Russian Federation
канд. физ.-мат. наук, доцент каф. дискретного анализа ЯрГУ, Международная лаборатория «Дискретная и вычислительная геометрия» им. Б. Н. Делоне,
Sovetskaya str., 14, Yaroslavl, 150000, Russia
H. Edelsbrunner
P. G. Demidov Yaroslavl State University
Russian Federation
Russian Federation
руководитель Международной лаборатории «Дискретная и вычислительная геометрия» им. Б. Н. Делоне,
Sovetskaya str., 14, Yaroslavl, 150000, Russia
References
1. Куваев Р. О., Кашин С. В., Капранов В. А., Эдельсбруннер Х., Мячин М. Л., Дунаева О. А., Русаков А. И. Новые компьютерные технологии эндоскопической диагностики в гастроэнтерологии и онкологии // Доказательная гастроэнтерология. 2013. Т. 2, № 1. С. 3–12. (Kuvayev R. O., Kashin S. V., Kapranov V. A., Edel’sbrunner Kh., Myachin M. L., Dunayeva O. A., Rusakov A. I. Novye komp’yuternye tekhnologii endoskopicheskoy diagnostiki v gastroenterologii i onkologii // Dokazatel’naya gastroenterologiya. 2013. Т. 2, № 1. S. 3–12. [in Russian])
2. Stehle T., Auer R., Gross S., Behrens A., Wulff J., Aach T., Winograd R., Trautwein C., Tischendorf J. Classification of colon polyps in NBI endoscopy using vascularization features // Medical Imaging 2009: Computer-Aided Diagnosis, eds. N. Karssemeijer and M. L. Giger, SPIE, 7260, 2009.
3. Munkres J. R. Elements of Algebraic Topology. Perseus, Cambridge, Massachusetts, 1984.
4. Edelsbrunner H. and Harer J. L. Computational Topology. An Introduction. Amer. Math. Soc., Providence, Rhode Island, 2010.
5. Tarjan R. E. Data Structures and Network Algorithms. SIAM, Philadelphia, Pennsylvania, 1983.
6. Felzenszwalb P. F., Huttenlocher D. P. Distance transforms of sampled functions // Theory Comput. 2012. V. 8. P. 415–428.
7. Freund Y. and Schapire R. E. A decision-theoretic generalization of on-line learning and an application of boosting // J. Comput. Sys. Sci. 1997. V. 55 P. 119–139.
Review
For citations:
Dunaeva O.A., Malkova D.B., Machin M.L., Edelsbrunner H. Segmentation of Clinical Endoscopic Images Based on the Classification of Topological Vector Features. Modeling and Analysis of Information Systems. 2013;20(6):162-173. (In Russ.) https://doi.org/10.18255/1818-1015-2013-6-162-173
Views:
1036
Email this article 