ANALYSIS OF SVM PARAMETRIZATION IN THE CLASSIFICATION OF MAMMOGRAPHIC TEXTURE IMAGES

ÁLVARO HENRIQUE DE ARAÚJO RUNGUE; ALEXEI  MANSO CORRÊA MACHADO; PEDRO   AUGUSTO PINHO FERRAZ; BERNARDO  AUGUSTO GODINHO DE OLIVEIRA; THIAGO  MELO MACHADO-COELHO2; WILLIAN  ANTÔNIO DOS SANTOS

doi:10.20906/CBA2022/697

ÁLVARO HENRIQUE DE ARAÚJO RUNGUE Pontifical Catholic University of Minas Gerais
ALEXEI MANSO CORRÊA MACHADO Pontifical Catholic University of Minas Gerais
PEDRO AUGUSTO PINHO FERRAZ Pontifical Catholic University of Minas Gerais
BERNARDO AUGUSTO GODINHO DE OLIVEIRA Pontifical Catholic University of Minas Gerais
THIAGO MELO MACHADO-COELHO2 FEDERAL UNIVERSITY OF MINAS GERAIS
WILLIAN ANTÔNIO DOS SANTOS Pontifical Catholic University of Minas Gerais

DOI: https://doi.org/10.20906/CBA2022/697

Keywords: Mammography, Image classification, Texture descriptors, Support vector machines

Abstract

Studies indicate that breast density is related to the risk of developing cancer since dense breast
tissue can hide lesions, causing cancer to be detected at later stages. In this paper we classication method using support vector machines (SVM) associated to data reduction techniques to classify mammographic texture. An analysis of the parameters that influence the efectiveness of texture classication is also provided. Experiments were conducted on a set of 4,000 mammographic exams from which regions of interest representing the most signicantly part of the texture of the breast tissue were extracted. Compared to other quantitative results found in the literature, the proposed multi-class SVM method using the radial basis function kernel and tuned parameters proved to be superior while classifying mammographic texture, reaching up to 99% of precision for 10% of recall.