A combined approach of multiscale texture analysis and interest point/corner detectors for microcalcifications diagnosis Annarita Fanizzi 1 , Teresa M.A. Basile 2,3 , Liliana Losurdo 1 , Roberto Bellotti 2,3 , Ubaldo Bottigli 4 , Vittorio Didonna 1 , Alfonso Fausto 5 , Raffaella Massafra 1 , Alfonso Monaco 3 , Marco Moschetta 6 , Pasquale Tamborra 1 , Sabina Tangaro 3 , and Daniele La Forgia 1 1 I.R.C.C.S. “Giovanni Paolo II” National Cancer Centre, Bari, Italy 2 Dept. of Physics, University of Bari “Aldo Moro”, Bari, Italy 3 INFN National Institute for Nuclear Physics, Bari Division, Bari, Italy 4 Dept. of Physical Sciences, Earth and Environment, University of Siena, Siena, Italy 5 Dept. of Diagnostic Imaging, University Hospital of Siena, Siena, Italy 6 Interdisciplinary Dept. of Medicine, University of Bari “Aldo Moro”, Bari, Italy Abstract. Screening programs use mammography as primary diagnos- tic tool for detecting breast cancer at an early stage. The diagnosis of some lesions, such as microcalcifications, is still difficult today for radi- ologists. In this paper, we proposed an automatic model for characteriz- ing and discriminating tissue in normal/abnormal and benign/malign in digital mammograms, as support tool for the radiologists. We trained a Random Forest classifier on some textural features extracted on a mul- tiscale image decomposition based on the Haar wavelet transform com- bined with the interest points and corners detected by using Speeded Up Robust Feature (SURF) and Minimum Eigenvalue Algorithm (MinEige- nAlg), respectively. We tested the proposed model on 172 ROIs extracted from 176 digital mammograms of the public database. The model pro- posed was high performing in the prediction of the normal/abnormal and benign/malignant ROIs, with a median AUC value of 98.46% and 94.19%, respectively. The experimental result was comparable with re- lated work performance. Keywords: Computer-Aided Diagnosis (CADx), Microcalcifications, Dig- ital Mammograms, Haar Wavelet Transform, SURF, Minimum Eigen Algorithm, Random Forest. 1 Introduction Breast cancer is the first cause of death among women all over the world. It is difficult to prevent it, but since the first studies [1, 2] it has been shown that an early diagnosis of breast lesions increases the chances of survival and reduce the mortality rate. Currently, screening programs use mammography [3–5] as primary diagnostic tool for detecting breast cancer at an early stage. Despite