Creating and sharing knowledge for telecommunications

Development of an Automation-Compatible Breast Cancer Imaging System using Microwaves

Felício, J.F. ; Bioucas-Dias, J. ; Costa, J.R. ; Fernandes, C. A.

Development of an Automation-Compatible Breast Cancer Imaging System using Microwaves, Proc IEEE Portuguese Meeting on Bioengineering ENBENG, Lisbon, Portugal, Vol. , pp. - , February, 2019.

Digital Object Identifier:

Abstract
Breast cancer is the most common cancer in women. Currently, mammography is the preferred screening modality for primary mass population screening. However, mammography requires the compression of the breast, and uses ionizing radiation, which is proven to be harmful to health in the long-term. There has been an effort to develop new imaging modalities that can be used for breast cancer detection as primary screening modality. One of the most promising is microwave imaging (MWI) [1]. It is documented that there is enough electromagnetic contrast between healthy and malignant tissues that may enable this technology [2].
In this framework, we developed a low-cost imaging system based on microwaves (MW). Contrarily to the big majority of MWI setups found in the literature, ours does not require any immersion liquid, making it more hygienic, practical and with lower maintenance. It consists of a set of antennas distributed around the breast that illuminate the tissues with MW energy.
The antennas collect the echoes from the body, from which we reconstruct the reflectivity of the breast, as to detect any abnormality. Moreover, we included a low-cost system for the reconstruction of the three-dimensional breast shape. These data are fed into our dedicated signal processing algorithms, in order to improve the imaging results. The algorithms are quite robust to the uncertainties in this kind of setups, and may be fully automated. We experimentally assessed the setup in our lab. To this end, we 3D-printed a realistic breast model with similar electromagnetic properties as an actual breast, including high-density
fibroglandular tissues. Even in low contrast scenario the results do show good detection of the tumor at the correct position, thus confirming the potential of the proposed setup for mass population breast cancer screening. The next step of this work is to approach a faster examination scenario, including more antennas, numerically more efficient and more robust signal processing algorithms, as to increase the resolution and sensitivity of the system.