Comparative study of feature classification methods for mass lesion recognition in digitized mammograms

Masala, Giovanni L. and Tangaro, Sonia and Golosio, B. and Oliva, P. and Stumbo, S. and Bellotti, R. and De Carlo, F. and Gargano, G. and Cascio, D. and Fauci, F. and Magro, R. and Raso, G. and Bottigli, U. and Chincarini, A. and De Mitri, I. and De Nunzio, G. and Gori, I. and Retico, A. and Cerello, P. and Cheran, S. C. and Fulcheri, C. and Lopez Torres, E. (2007) Comparative study of feature classification methods for mass lesion recognition in digitized mammograms. Il nuovo cimento C, 30 (3). pp. 305-316. ISSN 1826-9885

ncc9238.pdf - Published Version

Download (429kB) | Preview
Official URL:


In this work a comparison of different classification methods for the identification of mass lesions in digitized mammograms is performed. These methods, used in order to develop Computer Aided Detection (CAD) systems, have been implemented in the framework of the MAGIC-5 Collaboration. The system for identification of mass lesions is based on a three-step procedure: a) preprocessing and segmentation, b) region of interest (ROI) searching, c) feature extraction and classification. It was tested on a very large mammographic database (3369 mammographic images from 967 patients). Each ROI is characterized by eight features extracted from a co-occurrence matrix containing spatial statistics information on the ROI pixel grey tones. The reduction of false-positive cases is performed using a classification system. The classification systems we compared are: Multi Layer Perceptron (MLP), Probabilistic Neural Network (PNN), Radial Basis Function Network (RBF) and K-Nearest Neighbours classifiers (KNN). The results in terms of sensitivity (percentage of pathological ROIs correctly classified) and specificity (percentage of nonpathological ROIs correctly classified) are presented. MLP and RBF outperform other classification algorithms by about 8% of the area under the ROC curve.

Item Type: Article
Uncontrolled Keywords: Computer-aided diagnosis ; Digital imaging ; Image analysis ; Mammography
Subjects: 500 Scienze naturali e Matematica > 530 Fisica
Depositing User: Marina Spanti
Date Deposited: 20 Mar 2020 16:31
Last Modified: 20 Mar 2020 16:31

Actions (login required)

View Item View Item