**1. Introduction**

132 Imaging of the Breast – Technical Aspects and Clinical Implication

NHSBSP. National Health Service Breast Screening Programmes. (2003). Review of Radiation

NHSBSP. National Health Service Breast Screening Programmes. (2009). Commissioning and

Nishikawa R M. (2004). Advances in Breast Imaging: Physics, Technology, and Clinical

America, November 28–December 3, 2004, Chicago, USA. ISSN 0271-5333. Pedersen, K Nordanger J (2000). Mammografiscreening. Teknisk kvalitetskontroll – resultater

Peixoto, J. E. (2009). Controle de Qualidade em Mamografia. In Aguillar VLN, Bauab SP,

Pisano ED, Gatsonis C, Hendrick E, Yaffe M, Baum JK, Acharyya S, et al. 2005. Diagnostic

Rosenstein M, Andersen LW and Warner GG. (1985). Handbook of Glandular Tissue Doses in

Sobol WT and Wu X. 1997. Parametrization of mammography normalized average glandular

Stanton L, Villafana T, Day JL and Lightfoot DA. (1984). Dosage evaluation in mammography.

Vedantham S, Karellas A, Suryanarayanan S, Albagli D, Han S, Tkaczyk EJ, Landberg CE et al.

Wu X, Barnes G T, and Tucker D M. (1991). Spectral dependence of glandular tissue dose in

Wu X, Gingold E L, Barnes G T, and Tucker D M. (1994). Normalized average glandular dose

Yaffe MJ, Rowlands JA. (1997) X-ray detectors for digital radiography. Phys Med Biol;42:1–39. Young KC, Ramsdale ML and Bignall F. (1998). Review of dosimetric methods for

Young KC, Oduko JM, Bosmans H, Nijs K and Martinez L. (2006). Optimal beam quality selection in digital mammography. *Br J Radiol* 79: 981-990. ISSN: 1748-880X Zoetelief J, Veldkamp W J H, Thijssen M A O, and Jansen J T M. (2006). Glandularity and mean

screen-film mammography. *Radiology* 179:143–148. ISSN : 1527-1315

mammography. *Radiology* 193:83–89. ISSN : 1527-1315

(2000). Full Breast Digital Mammography with an Amorphous Silicon-based Flat Panel Detector: Physical Characteristics of a Clinical Prototype. Med Phys 27:558-567 Suryanarayanan S, Karellas A and Vedanthama S. (2004). Physical characteristics of a full-field

digital mammography system. *Nuclear Instruments and Methods in Physics Research A.*

in molybdenum target-rhodium filter and rhodium target-rhodium filter

mammography in the UK breast screening programme. *Radiat Prot Dosim*etry. 80:

glandular dose determined for individual women at four regional breast cancer screening units in the Netherlands. *Phys Med Biol*. 51:1807–1817. ISSN 1361-6560

Rowlands JA. (2002). The physics of computed radiography. Phys Med Biol;47: R123–R166. SEFM. Sociedad Española de Física Médica. (2008). Protocolo de Control de Calidad en

*Programmes*. ISBN 1 871997 99 2. Sheffield, UK

Norwegian Radiation Protection Authority (Norway).

Report 0604. (London, UK).

8537202444. Rio de Janeiro, Brazil

Health, Rockville, Maryland, USA

England Journal of Medicine. 353: 1-11.

*Radiology*.150: 577–584. ISSN :0033-8419

533:560–570. ISSN: 01689002

183–6. ISSN : 0144-8420

Mamografía Digital. Edicomplet (Madrid, Spain)

dose tables. *Med Phys* 24:547–554. ISSN: 0094-2405

Risk in Breast Screening. NHSBSP Publication No 54. *NHS Cancer Screening* 

Routine Testing of Full Field Digital Mammography Systems. NHSBSP Equipment

Applications. *The Radiological Society of North America* (RSNA 2004), Presented at the 90th Scientific Assembly and Annual Meeting of the Radiological Society of North

og evaluering etter fire års prøveprosjekt. StrålevernRapport 2000:2. Østerås:

Maranhão NM. Mama: diagnóstico por Imagem. *Revinter*, 2009. p. 83-106. ISBN-10:

Performance of Digital versus Film Mammography for Breast-Cancer Screening, New

Mammography, *HHS Publication FDA* 85-8239. Center for Devices and Radiological

To understand how we can optimize the spectrum of an X-ray beam to obtain the maximum contrast using the minimum dose for a given x-ray exam we need to know a little about how the X-rays are generated, how an x-ray tube works, how the energy spectrum of the output X-ray beam will look and how the X-rays interact with the human body and with the imager.

A general X-ray tube has two electrodes, the cathode and the anode, with a high voltage applied between them. Electrons are generated at the cathode. After the electrons are emitted from the cathode they are accelerated by the high voltage toward the anode (positive) electrode. The high speed electrons hitting the anode material then generate invisible radiation i.e. X-rays. The energy spectrum of the output X-rays is dependent on the anode-cathode voltage difference and on the material of the anode and is measured in electronvolts. The energy spectrum is a so called Brehmstrahlung radiation meaning that it is continuous over a wide range of energies and has more photons emitted at the lowest energies. The photon flux decreases to zero at the anode-cathode voltage difference level.

When the Brehmstrahlung radiation of the X-ray beam passes into the human body, some photons are absorbed while others pass through. The ones that have passed through are available for imaging. Because the absorption of the human body is higher at lower energy X-rays, the beam exiting the body has a higher average energy than the beam had when it entered the body. This effect is called beam hardening and it causes a decrease in the contrast of images and difficulties in CT reconstruction. Furthermore, the dose the patient receives from the very low energy X-rays serves no purpose; it does not contribute to the final image in any way.

Figure 1 demonstrates the difference between input and exit X-ray beams showing that the very low energy X-rays do not penetrate the human body at all and only add extra (unwanted) dose to the patient (Sutton 2009). Three cases are shown in the figure. The original input beam, an 80kVp X-ray beam from a W anode filtered through a 2 mm Al filter, is shown in blue and has the highest photon flux and widest energy spectrum. The red series shows the energy of the beam after having passed through simulated soft tissue (150 mm soft tissue and 50 mm water). This exit beam shows that only X-ray photons with >25keV energy will contribute to the image. The green series shows the energy of the beam

Contrast Enhancement in Mammography Imaging Including K Edge Filtering 135

photons for good imaging. Figure 2 shows the contrast/dose relationship for mammography. From the diagram, the optimum X-ray energy for imaging the breast is in

Fig. 2. The general relationship of contrast and dose to photon energy in mammography

There are several methods for limiting the input X-rays to an optimum range. They include dual energy imaging, monochromatic X-ray imaging, and quasi-monochromatic X-ray

In many cases X-ray images must capture both hard (bony) structures and soft tissues, which may be overlapping, in the same picture. A method to separate two different tissues, and optimize the contrast between them, is to acquire two images at different x-ray energies and do a weighted subtraction. This is called dual energy imaging. Taking images at low and high energies and doing a subtraction can remove either the soft or the bony structure

The data processing methods that make use of dual energy data are referred to as decomposition techniques and are divided into three classes based on the type of information returned by the technique. The three types of information about the object being imaged that may be returned by decomposition techniques are: (1) effective atomic number and density; (2) attenuation due to Compton scattering and electron density; or (3) the

physical densities of two known components (Walter, Wu et al. 2004).

2 This range will vary depending on the size and density of the breast.

the range of 18-21 keV (mean energy).2

(Sprawls).

imaging.

**3. Dual energy imaging** 

depending on the techniques used.

**3.1 Brief theory of dual energy imaging** 

after having passed through simulated tissue and bone (150mm soft tissue + 50 mm water + 30 mm bone). This exit beam now shows that only X-ray photons with >37 keV are available to contribute to the image. The very low energy input photons are completely absorbed in the body, adding extra patient dose, because there is no exit photon for imaging1.

Fig. 1. Comparison of the ratio of the incident photons (original beam) to the exit photons after going through different body tissues including water. Only the exit photons contribute to the imaging.

So, how can we optimize contrast of the x-ray image using the minimum dose?

## **2. Using the right x-ray energy range**

Figure 1 implies that the spectrum of the input x-ray beam has to be optimized for each imaging modality for minimum patient dose. For instance, a thicker body part such as the pelvis, which contains lots of bony structures, absorbs most of the low energy X-ray beams below 30 keV so we need to use higher X-ray beam energies to get a reasonable number of exiting x-ray photons for imaging while avoiding overdosing the patient with low energy Xrays which would be completely absorbed in the thick pelvis.

By contrast, imaging of the breast, which is smaller (thinner) relative to the pelvis and does not contain any bony structure, requires much lower X-ray energy to get sufficient X-ray

<sup>1</sup> Note that some low energy X-rays are removed by the inherent filtration of the X-ray tube. In many cases the exit window of the X-ray tube is made of 2mm Aluminum which removes most of the X-ray radiation below 15 keV.

photons for good imaging. Figure 2 shows the contrast/dose relationship for mammography. From the diagram, the optimum X-ray energy for imaging the breast is in the range of 18-21 keV (mean energy).2

Fig. 2. The general relationship of contrast and dose to photon energy in mammography (Sprawls).

There are several methods for limiting the input X-rays to an optimum range. They include dual energy imaging, monochromatic X-ray imaging, and quasi-monochromatic X-ray imaging.
