Characterization of breast nodules in digital images of ultrasonography, elastography and mammography using intelligent techniques

• Grant number: 12/24006-5
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: April 01, 2013
• End date: August 01, 2017
• Field of knowledge: Engineering - Biomedical Engineering - Bioengineering
• Principal Investigator: Homero Schiabel
• Grantee: Karem Daiane Marcomini
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated scholarship(s): 15/17302-5 - Creating an image database containing mammography, ultrasound and elastography for the segmentation of suspicious findings, BE.EP.DR

Abstract

Aiming to deal with the clinical worsening of breast cancer as well as to assist its early detection, many imaging diagnosis procedures have been developed or improved regarding both the image and the equipment. Modules of mammography CADx (Computer aided diagnosis) schemes have been proposed with the aim of providing diagnostic support, improving its accuracy. Among the available screening techniques, mammography by X-ray is dominant in the assessment of breast abnormalities. Ultrasound imaging uses to be the main complementary procedure with this intent. In order to reduce more significantly the number of unnecessary biopsies, elastography is a new technique of breast images acquisition, aiming also to increase the accuracy in mammographic findings interpretation. Thus, as part of a wider research concerning a CADx scheme encompassing the latest techniques for breast image acquisition and interpretation, this proposal intends to advance the detection and subsequent classification of suspicious findings, considering a comparative approach of the imaging by X-ray, ultrasound and elastography. The purpose is to set up a model for processing and classification of detected lesions, to be optimized according to the best features present in digital images obtained by those three types of technology. The model should take the use of processing techniques and investigate artificial intelligence algorithms to select the most efficient in the characterization of lesions according to the images set in each case.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

FLEURY, EDUARDO; MARCOMINI, KAREM. Performance of machine learning software to classify breast lesions using BI-RADS radiomic features on ultrasound images. EUROPEAN RADIOLOGY EXPERIMENTAL, v. 3, n. 1, p. 8-pg., 2019-08-05. (12/24006-5)

MARCOMINI, KAREM D.; FLEURY, EDUARDO F. C.; OLIVEIRA, VILMAR M.; CARNEIRO, ANTONIO A. O.; SCHIABEL, HOMERO; NISHIKAWA, ROBERT M.; ARMATO, SG; PETRICK, NA. Agreement between a computer-assisted tool and radiologists to classify lesions in breast elastography images. MEDICAL IMAGING 2017: COMPUTER-AIDED DIAGNOSIS, v. 10134, p. 8-pg., 2017-01-01. (12/24006-5, 15/17302-5)

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

MARCOMINI, Karem Daiane. Characterization of lesions in ultrasound and elastography images using machine learning methods. 2017. Doctoral Thesis - Universidade de São Paulo (USP). Escola de Engenharia de São Carlos (EESC/SBD) São Carlos.