Detection and evaluation of breast tumors on the basis of microcalcification analysis

Krzysztof Lewenstein , Krzysztof Urbaniak

Abstract

Long-term studies on the possibilities of aiding breast cancer diagnostics with the use of artificial neural natworks supported by promising results led the authors of this paper to take up another challenge, the aim of which was localization and classification of microcalcifications. The evaluation of mammographic images by a specialist is not easy, and often ambiguous. Microcalcifications possess ‘encoded’ significant diagnostic value while having a small size and low contrast. A large number of papers indicates that information included in mammogram can be the basis for extracting the features of microcalcifications and their satisfactory classification with the use of artificial neural networks [9, 10, 12, 13, 14, 15, 16, 19, 20]. Detection of microcalcifications is usually realized in two stages, i.e. by the analysis of mammographic image what results in defining microcalcifications concentration and then by trying to evaluate the degree of their malignance. The paper presents the results of research undertaken in order to evaluate mammograms in terms of detecting places of occurrence and evaluate the degree of their malignance. The evaluation took place on the basis of the analysis of mammographic image with the use of two types of neural networks: feed forward multi-layer MLBP networks and Fahlman networks.
Author Krzysztof Lewenstein IMIB
Krzysztof Lewenstein,,
- The Institute of Metrology and Biomedical Engineering
, Krzysztof Urbaniak ZFNSiPT
Krzysztof Urbaniak,,
- Department of Philosophy of Science, Sociology, and Fundamentals of Technology
Pages159-166
Publication size in sheets0.5
Book Jabłoński Ryszard, Březina Tomáš (eds.): Advanced Mechatronics Solutions, Advances in Intelligent Systems and Computing, 2016, Springer, ISBN 978-3-319-23921-7, 668 p., DOI:10.1007/978-3-319-23923-1
Keywords in Englishmicrocalcifications extraction method segmentation method binarization method Fahlman neural network MLBP neural network
DOIDOI:10.1007/978-3-319-23923-1_24
URL http://link.springer.com/chapter/10.1007/978-3-319-23923-1_24
Languageen angielski
Score (nominal)15
Citation count*0 (2016-08-26)
Cite
Share Share

Get link to the record
msginfo.png


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back