Breast cancer diagnosis using microwave

International Journal of Computer Science & Engineering Survey (IJCSES) Vol.5, No.3, June 2014
DOI : 10.5121/ijcses.2014.5304 41
BREAST CANCER DIAGNOSIS USING MICROWAVE
AND HYBRID IMAGING METHODS
Younis M. Abbosh
College of Electronics Engineering, Mosul University, Mosul, Iraq
ABSTRACT
This paper explains new imaging techniques that show promising results in breast cancer detection. The
presented techniques use microwave-based methods, wavelet analyses, and neural networks to get a
suitable resolution for the breast image. One of the presented techniques (hybrid method) uses a
combination of microwaves and acoustic signals to improve the detection capability. Some promising
results are shown and explained.
KEYWORDS
microwav imaging; breast cancer; hybrid imaging ; neural network; wavelet
1. INTRODUCTION
Breast cancer is one of the most common cancers diagnosed in women. In the western
countries, it is estimated that one in eleven women will develop breast cancer at some stage in
their life making it the highest life risk [1,2]. Early detection and timely medical treatment are key
factors affecting long-term survival of breast-cancer patients. Currently, the primary method for
breast screening is X-ray mammography and in some cases magnetic resonance imaging (MRI).
X-ray mammography is the golden diagnostic tool for breast cancer, but the technology has
several serious shortfalls. Firstly, it produces a relatively high false-negative rate, which can be as
large as 30% [3]. Secondly, screening mammography suffers from a high false-positive rate: on
average, 75% of breast biopsies prompted by a “suspicious” mammographic abnormality prove
benign [4]. Thirdly, screening mammography is less sensitive in women with radio-graphically
dense breast tissue prevalent in younger women, where it has been shown that X-ray
mammography has failed to detect up to 30% of cancers greater than 5 mm in diameter [5], due to
its relatively poor soft-tissue contrast.
Mammography’s other drawbacks include variability in radiological interpretation, and a
slight risk of inducing cancer due to the ionizing radiation exposure. Frequent monitoring is
difficult because of health concerns related to exposure to ionizing radiation [6]. Best prospects of
detecting malignant tissues using X-rays are when tumours reach the late stage of calcification,
which seriously jeopardises the success of treatment.
MRI is a highly sensitive imaging technique, but lacks information on specificity, and thus it
has high rate of false positive results [7]. The impact of misdiagnosis on the patient is
considerable, particularly when they are confronted with unnecessary surgeries. In general, MRI
can be considered to have a moderate success rate of correct diagnosis; however MRI has high
operational costs.
An ultimate diagnosis of all types of breast disease depends on a biopsy. A biopsy is an
invasive procedure to remove and examine tissue or cells for the presence of cancer. In most

42
cases the decision for a biopsy is based on mammography findings. Unfortunately, biopsy results
indicate that 80% of breast lesions detected by mammography are benign. This situation calls for
alternative diagnostic tools to reduce physical and mental suffering of patients caused by this false
positive diagnosis. Of particular interest is the development of low-cost diagnosis methods, which
could be easily accessed by as much people as possible.
Given the deficits of the current diagnostic tools, development of imaging modalities that
enhance, complement, or replace X-ray mammography or MRI has been a priority in many
countries and by many research groups around the globe. Some of the alternative modalities
under investigation by research groups around the world are ultrasound-based and microwave-
based methods.
Ultrasound imaging of the breast is capable of distinguishing between solid tumours and
fluid-filled cysts. Also, it can be used to evaluate lumps that are hard to see on a mammogram. As
ultrasound does not harm biological tissue, thus it can be applied frequently. This is of
importance, especially with respect to younger women for whom the risks from X-ray radiation
are most significant. However, ultrasound lacks spatial resolution, cannot image calcifications and
is very operator dependent. Breast cancer is relatively iso-echoic [8] (that is, the echo similarity of
two or more tissues as measured by ultrasonography) and hence, difficult to detect by ultrasound.
Thus, unnecessary biopsies are needed on benign masses [8].
In the last years, intensive research has been conducted at many universities on different
microwave-based imaging techniques for breast cancer detection. Some of promising techniques
use suitable parameters extracted from microwave signal collected at different places around
breast as input to neural network. These parameters are extracted from time and frequency
domain using wavelet transforms [9-11]. The breast model including the probes that are used to
collect the scattered signals is analyzed using the full-wave electromagnetic simulator CST
Microwave Studio. This research has enjoyed notable success especially in the design, and
development of different devices used in those techniques [12-20]. It has been demonstrated that
the use of microwave technology, combined with acoustic signals, properly designed microwave
devices can improve the resolution needed for early breast cancer detection. Those systems are
still under continuous improvements aiming at reaching a standard required to make the system
clinically reliable.
2. MICROWAVE IMAGING
The potential for using microwaves for detecting breast tumours is based on the concept of
tissue-dependent microwave scattering and absorption in the breast to exploit the contrast in the
dielectric properties of malignant and normal breast tissues. There are many approaches in which
microwaves can be utilized in the imaging tools as shown in Fig. 1.
It has been widely assumed that normal breast tissue is largely transparent to microwaves
because they are featured with a low relative permittivity and conductivity at the microwave
frequency bands, whereas lesions, which contain more water and blood are characterized by a
high relative permittivity and conductivity at the microwave frequencies and hence they cause a
significant backscatter [21]. Upon this assumption, which is supported by some measurements,
microwave imaging systems are being designed to detect the presence of a small object inside a
breast causing a considerably larger backscatter than the surrounding medium.
Microwave techniques involve the propagation of very low levels (1000 times less than a
mobile phone) of microwave energy through the breast tissue. The basis for tumour detection and
location is the difference in the electrical properties between normal and malignant breast tissue.

43
Normal breast tissue is largely transparent to microwave radiation, whereas malignant tissues,
which contain more water and blood, cause microwave signal backscattering. This scattered
signal can be picked by an array of microwave antennas and analysed using a computer.
Figure. 1 Microwave-based imaging methods
Figure. 2 Radar-based microwave imaging system of (a) monostatic, and (b) bistatic radars.
Figure. 3 Radar-based imaging of breast.

44
Figure. 4 Configuration of the radar system.
Figure. 5 Imaging result of the wideband radar system.
The radar approach to microwave imaging (Fig. 2) employs generating and receiving
short pulses for various locations of probe antenna or alternatively by an array antenna. The
processed signals for various locations of a probe antenna or from array elements are combined to
form a two or three-dimensional image showing the location of a highly reflecting object
representing a cancerous tissue [12].
The configuration, shown in Fig. 2(a), is based on the principle of monostatic radar. In
this configuration, the same antenna is used for both transmitting and receiving of a microwave
signal. The configuration shown in Fig. 2(b) uses two antennas, which are displaced by a certain
distance. In this case, the microwave imaging system is based on the principle of bistatic radar.
When using the radar-based techniques to image human breast, the procedure is as shown
in Fig. 3. The configuration of a prototype radar system is shown in the Fig. 4 [16]. The system
consists of a circular cylindrical scanning platform with a resolution of 1° to support a breast
phantom, and a mechanical scanning platform with resolution of 0.1 mm in the vertical axis. The
scanning platform supports an array of wideband antennas [22]. The antenna is connected to a
microwave Vector Network Analyzer that measures the scattered signals. The collected scattered
signals are then processed in a personal computer to get an image of the breast.
The imaging capabilities of the aforementioned radar system are carried out for an artificial
breast phantom. Sample of the obtained images is shown in Fig. 5. The figure clearly shows the
boundaries of the layer representing the skin, size and shape of the object that represents the
tumor.

45
The microwave-based imaging systems face two serious challenges before being ready for
clinical tests. First, the success of those systems depends on a substantial contrast in the dielectric
properties of malignant and normal breast tissues. A large scale study to experimentally determine
the dielectric properties of a variety of normal, malignant and benign breast tissues, measured at
the microwave range has recently been conducted in the USA [23, 24]. It has been demonstrated
that the dielectric contrast varies widely from individual to individual and with age and some
other biological factors. It has also been found that although the contrast between malignant and
normal adipose-dominated tissues in the breast is considerable, the contrast in the dielectric
properties is no more than about 10% between malignant and normal glandular/fibro-connective
tissues in the breast [24]. By using the scattering theory of electromagnetic fields [25], it is
possible to show that when the dielectric contrast between the normal and malignant tissue is low,
resolution of the microwave imaging system deteriorates rapidly resulting in a blurry image of the
breast.
The second important challenge that limits the success of the microwave-based imaging
systems originates from the nature of the microwave signals. The limited resolution comes from
the fact that the heterogeneous structure of the breast causes multiple scatterings and reflections
for the microwave signal while penetrating inside the tissues. This increases the uncertainty in
estimating the three-dimensional image of the breast. That uncertainty can only be removed by
additional information about the breast tissues.
3. HYBRID IMAGING
In order to remove the obstacles facing the success of the microwave-based imaging methods, the
hybrid technique is proposed and tested [17,19,20]. The hybrid technique utilizes the dielectric
(electrical properties) and elasticity (mechanical properties) contrast between tumors and healthy
tissue in order to produce a three dimensional image of the breast. The information from the
hybrid image significantly enhances early diagnostic accuracy.
The mechanical properties of biological tissues are important indicators for biomedical
diagnosis since they are generally correlated with tissue pathological changes [26]. For example,
different kinds of carcinomas of the breast have been found to be harder than its surrounding
normal tissues [27]. Although different in terms of their elasticity, some tumors are not readily
detectable by conventional imaging modalities, especially in the presence of complex background
alterations such as scar tissue, or other benign phenomena [28].
A schematic diagram of the system is shown in Fig. 6 [17-20]. Under operator guided
computer control, a very short frequency-modulated pulse is applied from an ultra wideband
(UWB) antenna beside the breast of a patient in the face-down (prone) position via a switching
unit. At the same time, an acoustic signal is applied from a transducer below the breast. The
proposed hybrid system utilizes the combined benefits of microwaves, with UWB spectrum, and
acoustic excitations to produce a three-dimensional image. In this method, microwaves are
employed to give a full view of the dielectric contrasts, while acoustics give the elasticity
distributions within the breast. The two distributions are then used to produce a final image with
high contrast and resolution.
The scattered signals, which are a consequence of the difference in the electrical and
mechanical properties of different types of breast tissues, are collected by a cylindrical array of
UWB antennas encircling the breast. Parameters of the scattered signals, i.e. amplitude, time
delay, frequency shift and their angular distribution, correlate with characteristics of the breast
tissue. For example, shift in frequency, which is called Doppler frequency, is due to elasticity
contrast, whereas time delay is due to dielectric contrast. The detection unit shown in Fig. 6

46
captures those parameters, and the signal conditioning unit is designed to digitize and improve the
quality of the data for storage on a high speed computer. Operation of the different modules of
the system are controlled from a computer, which would also be used for calibration, data
collection, processing, and to display the images. Certain processing algorithms are used to
produce a clear three-dimensional image of the breast. The operator is able to distinguish tumors
from the healthy tissues of the breast by identifying the different breast regions characterized by
high backscatter and high stiffness.
The switching unit depicted in Fig. 6 includes a matrix of directional couplers and power
dividers. The unit is used to direct the generated microwave pulse towards the transmitting
antenna and the scattered pulses towards the detection unit and achieve a high level of isolation
between them. The detection unit includes a matrix of correlators, which have the ability to detect
the amplitude and phase variations of the received signals.
In order to remove undesired reflections due to the mismatch between the antenna, the skin
layer and the space separating them, the breast under test is immersed in a coupling liquid which
has material properties that reduce the backward scattering at the skin layer, and thus increasing
the dynamic range of the system [29].
Snapshots from the results of imaging using the hybrid technique are shown in Fig. 7. The
results in that figure reveal the possibility of detection as small as 1 mm tumors. This is a
promising result. However, further work needs to be done so that the imaging can be performed in
more realistic environments using heterogeneous breast phantoms.
Figure. 6 A schematic diagram showing the hybrid imaging system [14], [16], [17].

47
Figure. 7 The successful detection of small tumors using the hybrid techniques on artificial breast phantom.
REFERENCES
[1] Australian Institute of Health and Welfare & Australasian Association of Cancer Registries (AACR),
“Cancer in Australia 2001” AIHW Cat. # CAN23, Canberra: AIHW (Cancer Series no. 28), pp64-68,
2004
[2] Breast cancer network Australia, http://www.bcna.org.au/content/view/526/588/, 2008.
[3] P.Huynh, A. Jarolimek, and S. Daye, “The false-negative mammogram,” Radiographics, vol. 18, no.
5, pp. 1137–1154, 1998.
[4] S.Nass, I. Henderson, and J. Lashof, Mammography and beyond: Developing technologies for the
early detection of breast cancer, National Academy Press, Washington D.C., 2001.
[5] E.A.Sickles, “Nonpalpable, circumscribed, noncalcified, solid breast masses: Likelihood of
malignancy based on lesion size and age of patient,” Radiology, vol. 192, pp. 439–442, 1994.
[6] E.Fear, P. Meaney, and M. Stuchly, “Microwaves for breast cancer detection” IEEE Potentials, vol.
22, no. 1, pp. 12-18, 2003.
[7] N.Hoogerbrugge, et al, “The impact of a false-positive MRI on the choice for mastectomy in BRCA
mutation carriers is limited,” Annals of Oncology, vol. 19, no. 4, pp. 655-659, 2008.
[8] D.Melodelima, J. Bamber, F. Duck, J. Shipley, and L. Xu, “Elastography for breast cancer diagnosis
using radiation force: system development and performance evaluation,” Ultrasound in Medicine and
Biology, vol. 32, no. 3, pp. 387–396, 2006.
[9] Younis M.Abbosh, Ammar F. Yahya, , and Amin Abbosh, “Neural Networks for the Detection and
Localization of Breast Cancer,” International Conference on Communications and Information
Technology (ICCIT), pp 156-159, March 2011
[10] Ammar F.Yahya, Younis M. Abbosh, and Amin Abbosh, “Microwave Imaging Method Employing
Wavelet Transform and Neural Networks for Breast Cancer Detection,” Proceedings of the Asia-
Pacific Microwave Conference 2011, pp 1418-1421
[11] A.Abbosh, Y. Abbosh, and A. Yahya “Modeling and Imaging of Breast Using the Combined Effect of
Microwave and Acoustic Pulses,” Proceedings of the Asia-Pacific Microwave Conference 2011, pp
1790-1793
[12] W.Khor, M. Bialkowski, A. Abbosh, N. Seman, and S. Crozier, “An ultra wideband microwave
imaging system for breast cancer detection,” IEICE Trans.Communications, vol. E-90B, no. 9, pp.
2376-2381, 2007.
[13] A.Abbosh, K. Kan, and M. Bialkowski, “A compact UWB planar tapered slot antenna for use in a
microwave imaging system,” Microwave and Optical Technology Letters, vol. 48, no. 11, pp. 2212-
2216, 2006.
[14] W.Khor, H. Wang, M. Bialkowski, A. Abbosh, and N. Seman, “An experimental and theoretical
investigation into capabilities of a UWB microwave imaging radar system to detect breast cancer,”
European Microwave Conference, 2007.
[15] A.Abbosh, M. Bialkowski, and S. Crozier, “A simple model for electromagnetic scattering due to
breast tumour,” IEEE-Antennas and Propagation Symposium, 2008.

48
[16] W.C. Khor and M.E. Bialkowski, “Investigations into Cylindrical and Planar Configurations of a
Microwave Imaging System for Breast Cancer Detection” Proc. IEEE Antennas and Propagation
Society International Symposium, pp. 263-266, Albuquerque, New Mexico, USA, Jul 2006.
[17] A.Abbosh, S. Crozier, "Strain Imaging of the Breast by Compression Microwave Imaging," Antennas
and Wireless Propagation Letters, IEEE , vol.9, no., pp.1229-1232, 2010
[18] A.Abbosh, and S. Crozier, "Hybrid Imaging Method for Early Breast Cancer Detection," Biomedical
Engineering Conference, 2008. CIBEC 2008. Cairo International , vol., no., pp.1-4, 18-20 Dec. 2008
[19] A.Abbosh, "Strain imaging of breast using ultra-wideband pulse," Microwave Conference
Proceedings (APMC), 2010 Asia-Pacific , vol.,
[20] A.Abbosh, "Early breast cancer detection using hybrid imaging modality," Antennas and Propagation
Society International Symposium, 2009. APSURSI '09. IEEE , vol., no., pp.1-4, 1-5 June 2009
[21] Y.Xie, B. Guo, L. Xu, J. Li, and P. Stoica, “Multistatic adaptive microwave imaging for early breast
cancer detection,” IEEE Trans. Biomed. Eng., vol. 53, pp. 1647–57, 2006.
[22] A.Abbosh, H.K. Kan and M.E. Bialkowski “Design of compact directive ultra wideband antipodal
antenna” Microwave and Optical Techn. Letters. vol. 48, No. 12, pp. 2448-2450, 2006.
[23] M.Lazebnik, et al, “A large-scale study of the ultrawideband microwave dielectric properties of
normal breast tissue obtained from reduction surgeries,” Physics in Medicine and Biology, vol. 52, pp.
2637–2656, 2007.
[24] M.Lazebnik, et al, “A large-scale study of the ultrawideband microwave dielectric properties of
normal, benign and malignant breast tissues obtained from cancer surgeries,” Physics in Medicine and
Biology, vol. 52, pp. 6093–6115, 2007.
[25] D.Fotiadis, and C. Massalas, Mathematical Methods in Scattering Theory and Biomedical
Engineering, World Scientific, 2007.
[26] Y.Fung, Biomechanics: Mechanical properties of living tissues, 2nd ed., Springer-Verlag, New York,
1993.
[27] W.Donegan, J. Spratt, and J. Orsini, Cancer of the breast, 5th ed., W.B. Saunders, New York, 1987.
[28] L. Newcomer, et al, ‘‘Detection method and breast carcinoma histology,’’ Cancer, vol. 95, pp. 470–
477, 2002.
[29] A. Abbosh, M. Bialkowski, and S. Crozier, “Investigations into optimum characteristics for the
coupling medium in UWB breast cancer imaging systems”, IEEE-Antennas and Propagation
Symposium, USA, 2008.
Author
Younis M. Abbosh born in 1957 in Mosul-IRAQ. He awarded BSc in Electronic and
Communication Engineering from University of Mosul, Mosul- IRAQ in 1979.Next, he
awarded the MSc in Electronic and Communication from University of Mosul in 1985,
and PhD in Electronic and Communication in 2006 from same University. From 2007-
2010, Dr Younis worked as assistance dean of College of Electronics Engineering at
Mosul University. From 2013-now he is working as Head of Computer and Information
Department in College of Electronics Engineering.
Through his academic life he published over 10 papers in field of Electronics, Communications, and Digital
Signal Processing.

Breast cancer diagnosis using microwave

More Related Content

What's hot

Similar to Breast cancer diagnosis using microwave

Recently uploaded

Breast cancer diagnosis using microwave