Computer modeling to detection of tumors in mammography by images digital subtraction

• Grant number: 12/21825-5
• Support Opportunities: Regular Research Grants
• Start date: March 01, 2013
• End date: May 31, 2015
• Field of knowledge: Engineering - Biomedical Engineering - Bioengineering
• Principal Investigator: Homero Schiabel
• Grantee: Homero Schiabel
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

This work aims to propose a model involving digital subtraction of images aquired at different X-ray energy levels. The purpose is to allow the detection of malignant breast lesions which, in the conventional exam form, would be totally hidden when superimposed by tissues of similar absorption characteristics. This research is addressed to dense breast cases, which are commonly poor in contrast due to the presence of fibroglandular high density tissue. In order to provide this investigation, the images will be yielded by computer simulation of the main breast tissues - fat, glandular and the carcinoma. The behavior of gray scale variations in mammographic images has to be observed from the respective tissues absorption coefficients, taken with different thicknesses, and submitted to different energy values in the typical mammography exam range. The hypothesis is that differential absorption variations happen between the normal structures and the lesions inside the breast. Results from the computer simulation should be evaluated based on actual images acquired from properly defined exposures of breast phantoms in conventional and in digital mammography equipment. (AU)