Screening programs use mammography as primary diagnos- tic tool for detecting breast cancer at an early stage. The diagnosis of some lesions, such as microcalcications, is still dicult 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 classier 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.
Scheda prodotto non validato
Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo
|Titolo:||A Combined Approach of Multiscale Texture Analysis and Interest Point/Corner Detectors for Microcalcifications Diagnosis|
|Data di pubblicazione:||2018|
|Appare nelle tipologie:||2.1 Contributo in volume (Capitolo o Saggio)|