**5. Dosimetry in mammography**

One of the pillars underpinning the analysis of the risks-benefits of mammography is the accurate knowledge of the imparted doses, since it is well-established that there is an association between breast dose and the increased incidence of breast cancer. Assessments of breast doses are particularly important in breast screening programmes in which large groups of asymptomatic women undergo mammographic examinations. As in other radiological examinations, dose values are indicative of the diagnostic adequacy of the mammography technique selected in clinical practice. In addition, knowledge of dose values is essential for optimisation strategies developed to minimise doses while maintaining the necessary image quality.

The X-ray spectrum in mammography is of low energy and the depth dose within the breast decreases rapidly. Due to this, it is important to use a dosimetric quantity which gives a measure of the dose to the whole organ. Glandular tissue is the most vulnerable in the breast as compared to adipose, skin and areolar (nipple) tissues (Hammerstein et al., 1979). At present, it is widely accepted that mean glandular dose (DG) is the most appropriate dosimetric quantity to predict the risk of radiation carcinogenesis. Therefore, this quantity has been recommended by several national and international organisations (NCRP, 1986; IPSM, 1989; IAEA, 2007) and it is the quantity used in many national protocols for mammographic quality assessment (CEC, 1996, 2006; ACR, 1999; IAEA, 2007). The factors that affect DG are the X-ray beam quality and breast thickness and composition. These two latter parameters have a larger variability than the former, varying both within and between populations, and the latter with women's age as well. Even when the average glandularity would be the same, its distribution is unpredictable and changes from breast to breast.

Direct measurements of DG are not possible for individual breasts and, therefore, DG is derived from the entrance surface dose (or a related quantity) using adequate conversion factors (ICRU, 2005; NCRP, 2004). These factors were initially measured (Hamerstein, 1979; Stanton, 1984) and further calculated by means of Monte Carlo techniques (Rosenstein, 1980; Dance, 1990, 2000, 2009; Wu et al, 1991, 1994; Klein et al., 1997; Boone et al., 2002). This latter approach allows for the possibility of estimating conversion factors for a wide range of input spectra and breast features. Differences among the conversion factors (cG) obtained by several authors mainly arise from differences in breast model geometry, mammographic X-

Artefacts are undesirable characteristics which are not related to the mammary anatomic structures of a radiographic image. They can hinder the image by hiding or simulate a lesion

Artefacts can be caused by the source of X-rays, the beam filter, the compression device, breast support, grid, and flaws in processing, amongst others. In digital mammography, besides the sources just cited, the non-uniformity in the response of the elemental detectors may also generate artefacts, owing to the results of an inadequate flat-fielding. Another drawback in the digital system is the presence of reminiscent images (ghost images), resulting from previous exposures (ICRU, 2009). The latest appears more often with CR

One of the pillars underpinning the analysis of the risks-benefits of mammography is the accurate knowledge of the imparted doses, since it is well-established that there is an association between breast dose and the increased incidence of breast cancer. Assessments of breast doses are particularly important in breast screening programmes in which large groups of asymptomatic women undergo mammographic examinations. As in other radiological examinations, dose values are indicative of the diagnostic adequacy of the mammography technique selected in clinical practice. In addition, knowledge of dose values is essential for optimisation strategies developed to minimise doses while maintaining the

The X-ray spectrum in mammography is of low energy and the depth dose within the breast decreases rapidly. Due to this, it is important to use a dosimetric quantity which gives a measure of the dose to the whole organ. Glandular tissue is the most vulnerable in the breast as compared to adipose, skin and areolar (nipple) tissues (Hammerstein et al., 1979). At present, it is widely accepted that mean glandular dose (DG) is the most appropriate dosimetric quantity to predict the risk of radiation carcinogenesis. Therefore, this quantity has been recommended by several national and international organisations (NCRP, 1986; IPSM, 1989; IAEA, 2007) and it is the quantity used in many national protocols for mammographic quality assessment (CEC, 1996, 2006; ACR, 1999; IAEA, 2007). The factors that affect DG are the X-ray beam quality and breast thickness and composition. These two latter parameters have a larger variability than the former, varying both within and between populations, and the latter with women's age as well. Even when the average glandularity would be the same, its distribution is unpredictable and changes from breast to breast.

Direct measurements of DG are not possible for individual breasts and, therefore, DG is derived from the entrance surface dose (or a related quantity) using adequate conversion factors (ICRU, 2005; NCRP, 2004). These factors were initially measured (Hamerstein, 1979; Stanton, 1984) and further calculated by means of Monte Carlo techniques (Rosenstein, 1980; Dance, 1990, 2000, 2009; Wu et al, 1991, 1994; Klein et al., 1997; Boone et al., 2002). This latter approach allows for the possibility of estimating conversion factors for a wide range of input spectra and breast features. Differences among the conversion factors (cG) obtained by several authors mainly arise from differences in breast model geometry, mammographic X-

**4.4 Artefacts** 

on detection.

systems or aSe based flat panel detectors.

**5. Dosimetry in mammography** 

necessary image quality.

ray spectral data, photon interaction cross-sections and Monte Carlo codes. Other important factors affecting the calculations are the mammography system's characteristics and the imaging system components and geometry. The differences in cG values quoted by different authors were as high as 15-16% (Klein, 1997; ICRU, 2005; Dance 2000).

The breast model most commonly adopted (Fig. 8) has a central region consisting of a homogeneous mixture of adipose and glandular tissue surrounded by a layer at all sides, except for the one corresponding to the chest wall representing the skin. It is assumed that the breast is firmly compressed by a polycarbonate compression paddle. The percentage of breast glandularity is defined as the fraction by weight of glandular tissue at the central region (without skin). Most authors employ the elemental tissue composition published by Hammerstein et al. (1979).

Initially, it was assumed that a 50:50 mixture of adipose and glandular tissues was representative of a typical breast (Hammerstein, 1979). On this basis, phantoms of several thicknesses were constructed assuming this "standard" composition with the aim of facilitating DG estimates in the practice. This assumption implied that the fraction of glandular tissue was independent of compressed breast thickness. On the basis of this data, the standard "phantom" was defined as a 4.5 cm thickness of PMMA, representing the "standard breast" (4 cm thick and 50%/50% glandular/adipose tissue) (IPEM, 1989; CEC, 1996). Data indicating that the composition of the average compressed breast deviates from the 50:50 composition has been published (Geise, 1996; Klein et al., 1997; Young et al., 1998; Chevalier, 1998; Beckett, 2000; Zoetelief et al., 2006). In addition, it was found that breast glandularity decreases when the compressed breast thickness increases. In some of these works it is also determined that the equivalent thickness of PMMA gives the same incident air kerma at its upper surface as for that of a breast of a specified thickness and composition (Geise, 1996; Dance et al., 2000, 2009; Kruger, 2001; Argo, 2004).

Fig. 8. Breast model geometry. The rectangular section represents a vertical cross-section through the breast coplanar with the focal spot of the X-ray tube. The D-shape section represents the breast in craniocaudal projection. The shaded and outer regions represent, respectively, the breast parenchyma and the skin (0.5 cm of adipose tissue). In the work of Dance (2009) it is used as a voxelised breast model.

#### **5.1 Practical issues**

DG is generally calculated through the following relationship (ICRU, 2005):

$$D\_G = \mathbf{c}\_G \cdot \mathbf{K}\_{a,i} \tag{1}$$

Image Quality Requirements for Digital Mammography in Breast Cancer Screening 127

In the ACR approach (ACR, 1999), *Ka,i* is directly measured by placing the ionisation chamber adjacent to the ACR phantom at the level of the entrance surface of the phantom. The chamber is positioned at 4 cm from the chest wall. The compressor plate is located above the phantom and the chamber. The exposure conditions are those used clinically for a

Ka,i can be also measured using TLD dosimeters (ACR, 1999; CEC, 1996) placed on the entrance surface of the phantom or breast. The TLDs have to be calibrated in terms of air kerma free-in-air against a suitable ionisation chamber and dosimeter. Hence, the entrance dose measured by TLDs placed on the phantom or patient surface includes backscatter. Measurements performed with TLD dosimeters are influenced by many factors, including the performance of the instrument and those related to procedure of dosimeter preparation and handling. In addition, TLDs' response dependence of scatter gives rise to underestimations of *Ka,i* in a magnitude that is dependent on the dosimeter's thickness and the relative amount of backscatter radiation (Dance at al., 1999). Another factor that limits the use of TLD dosimeters is related with its visibility in the breast image. It is recommended that they be positioned on the upper inner quadrant of the breast so as to

The ACR protocol adopted the cG values calculated by Wu (1991, 1994; Sobol, 1997). The European and the IAEA protocols (CEC, 1996, 2006; IAEA, 2007) recommended the use of the cG values from Dance (1990; 2000; 2009). Fig. 8 and Table 1 summarise the details used by both authors to perform the Monte Carlo calculations. The cG values depend on the beam quality (half value layer (HVL)), breast thickness and breast composition. It is important to measure HLV and compressed breast thickness with accuracy in order to minimise the errors in the DG estimate. Narrow beam geometry is recommended for HVL measurements with the aim of reducing the influence of scattered radiation (IPEM, 2005; IAEA, 2007). In addition, Al filters of high purity (>99%) should be used and the compressor plate should be in place during the measurements. Errors in the compressed breast thickness measurement are due to the compressor plate which can bend and deform considerably. Several authors have proposed methods to gain accuracy in these measurements (Burch, A, 1995; Maria S.

DG values derived from phantom measurements are useful for 1) simplifying the follow-up of the mammography system's performance, 2) comparing with references or limiting values allowing the checking of the compliance of the equipment with recommendations, 3) checking if the exposure factors selected by the mammography system are suitable in terms

The evaluation of the mean glandular dose with large patient's samples enables a more direct evaluation of the risk of radiation induced cancer. However, it is difficult to know the composition of individual breasts needed to determine the conversion factors. A fairly widespread method is to determine the composition of the breast from its image. To avoid bias or subjective criteria the BI-RADS criteria are used (ACR, 1998) so that the individual breasts are classified by the radiologist into one of four possible groups according to its

of radiation dose, 4) for developing optimisation strategies.

ascribed glandularity (0%, 25%, 75% and 100%).

4 cm compressed breast.

Nogueira., 2011).

minimise interference with breast tissues. **5.1.1.1 Mean glandular dose (DG) estimates** 

where *Ka,i* represents the incident air kerma (without backscatter) and cG is the appropriate conversion factor. The incident air kerma is the air kerma free in the air (without backscatter) at the central axis of the incident X-ray beam at the skin-entrance plane which yields the desired image optical density (screen-film mammography) or signal:noise ratio (digital mammography).

#### **5.1.1 Determination of the incident air kerma**

Two approaches have been used to determine the incident air kerma, Ka,i. In the European and IAEA approach (CEC, 1996, 2006; IAEA, 2007), this quantity is calculated from the measured value of the X-ray tube output, Y(d), in terms of air kerma per tube-currentexposure-time product (mGy/mAs), measured at a distance d from the focal spot of the mammography unit. The value of Ka,i at the focus to surface distance for the phantom or for compressed breast (d') is determined as follows:

$$K\_{a,i} = Y(d)P\_{it} \cdot \left(\frac{d}{d}\right)^2\tag{2}$$

Where *Pit* is the mAs employed for a given exposure of the compressed breast or phantom, which is determined from the AEC post-exposure readout. The tube output has to be measured using an ionisation chamber with a flat energy response (CEC, 1996; ICRU, 2009) and conveniently calibrated for the mammography beam qualities. The ionisation chamber is placed at 6 cm from the chest wall and laterally centre. The distance d is usually fixed at 4.5 cm above the breast support. The compression plate should be about 10 cm above the chamber so as to avoid backscatter effects.


Table 1. Important parameters considered by Dance and Wu for the calculation of the conversion factors using Monte Carlo techniques.

In the ACR approach (ACR, 1999), *Ka,i* is directly measured by placing the ionisation chamber adjacent to the ACR phantom at the level of the entrance surface of the phantom. The chamber is positioned at 4 cm from the chest wall. The compressor plate is located above the phantom and the chamber. The exposure conditions are those used clinically for a 4 cm compressed breast.

Ka,i can be also measured using TLD dosimeters (ACR, 1999; CEC, 1996) placed on the entrance surface of the phantom or breast. The TLDs have to be calibrated in terms of air kerma free-in-air against a suitable ionisation chamber and dosimeter. Hence, the entrance dose measured by TLDs placed on the phantom or patient surface includes backscatter. Measurements performed with TLD dosimeters are influenced by many factors, including the performance of the instrument and those related to procedure of dosimeter preparation and handling. In addition, TLDs' response dependence of scatter gives rise to underestimations of *Ka,i* in a magnitude that is dependent on the dosimeter's thickness and the relative amount of backscatter radiation (Dance at al., 1999). Another factor that limits the use of TLD dosimeters is related with its visibility in the breast image. It is recommended that they be positioned on the upper inner quadrant of the breast so as to minimise interference with breast tissues.

#### **5.1.1.1 Mean glandular dose (DG) estimates**

126 Imaging of the Breast – Technical Aspects and Clinical Implication

where *Ka,i* represents the incident air kerma (without backscatter) and cG is the appropriate conversion factor. The incident air kerma is the air kerma free in the air (without backscatter) at the central axis of the incident X-ray beam at the skin-entrance plane which yields the desired image optical density (screen-film mammography) or signal:noise ratio

Two approaches have been used to determine the incident air kerma, Ka,i. In the European and IAEA approach (CEC, 1996, 2006; IAEA, 2007), this quantity is calculated from the measured value of the X-ray tube output, Y(d), in terms of air kerma per tube-currentexposure-time product (mGy/mAs), measured at a distance d from the focal spot of the mammography unit. The value of Ka,i at the focus to surface distance for the phantom or for

> , ' ( ) *a i it <sup>d</sup> K YdP*

Where *Pit* is the mAs employed for a given exposure of the compressed breast or phantom, which is determined from the AEC post-exposure readout. The tube output has to be measured using an ionisation chamber with a flat energy response (CEC, 1996; ICRU, 2009) and conveniently calibrated for the mammography beam qualities. The ionisation chamber is placed at 6 cm from the chest wall and laterally centre. The distance d is usually fixed at 4.5 cm above the breast support. The compression plate should be about 10 cm above the

cG DGN D N G pg gcs gcs Units mrad/R mrad/R mGy/mGy mGy/mGy mGy/mGy

(cm) 3 – 8 2 – 8 2 – 11 2 – 11

(kV) 10 – 35 23 – 50 25, 26, 28, 30, 32 25 - 40

Grid ----- ----- In place In place Image receptor ----- Yes (screen) Yes (screen) Yes (screen)

distance 60 cm 65

Table 1. Important parameters considered by Dance and Wu for the calculation of the

Rh/Rh

Compressor In place compressing the breast

2

Wu, 1991 Wu, 1994 Dance, 1990 Dance, 2000 Dance, 2009

Homog. mixture. 50%

Mo/Mo; W/Mo; W/Rh; W/Pd; W/Al

*d*

(digital mammography).

**5.1.1 Determination of the incident air kerma** 

compressed breast (d') is determined as follows:

chamber so as to avoid backscatter effects.

Spectra Mo/ Mo Mo/Rh;

conversion factors using Monte Carlo techniques.

Homogeneous mixture. 100%; 50%; 0%

Breast composition (% glandular tissue)

Breast thickness

Tube voltage

Source image

*D cK G G ai* , (1)

(2)

Homog. mixture 0% - 100%

Mo/Mo; Mo/Rh Rh/Rh; W/Rh

Homog. mixture 0% - 100%

> W/Ag W/Al

The ACR protocol adopted the cG values calculated by Wu (1991, 1994; Sobol, 1997). The European and the IAEA protocols (CEC, 1996, 2006; IAEA, 2007) recommended the use of the cG values from Dance (1990; 2000; 2009). Fig. 8 and Table 1 summarise the details used by both authors to perform the Monte Carlo calculations. The cG values depend on the beam quality (half value layer (HVL)), breast thickness and breast composition. It is important to measure HLV and compressed breast thickness with accuracy in order to minimise the errors in the DG estimate. Narrow beam geometry is recommended for HVL measurements with the aim of reducing the influence of scattered radiation (IPEM, 2005; IAEA, 2007). In addition, Al filters of high purity (>99%) should be used and the compressor plate should be in place during the measurements. Errors in the compressed breast thickness measurement are due to the compressor plate which can bend and deform considerably. Several authors have proposed methods to gain accuracy in these measurements (Burch, A, 1995; Maria S. Nogueira., 2011).

DG values derived from phantom measurements are useful for 1) simplifying the follow-up of the mammography system's performance, 2) comparing with references or limiting values allowing the checking of the compliance of the equipment with recommendations, 3) checking if the exposure factors selected by the mammography system are suitable in terms of radiation dose, 4) for developing optimisation strategies.

The evaluation of the mean glandular dose with large patient's samples enables a more direct evaluation of the risk of radiation induced cancer. However, it is difficult to know the composition of individual breasts needed to determine the conversion factors. A fairly widespread method is to determine the composition of the breast from its image. To avoid bias or subjective criteria the BI-RADS criteria are used (ACR, 1998) so that the individual breasts are classified by the radiologist into one of four possible groups according to its ascribed glandularity (0%, 25%, 75% and 100%).

Image Quality Requirements for Digital Mammography in Breast Cancer Screening 129

The method followed for patient dosimetry relies on the results obtained in two studies (Young et al 1998; Beckett et al., 2000) that have each independently estimated the breast composition of women attending for screening. As a result, the average breast composition as a function of breast thickness was established for two age groups. One age group (aged 50 to 64) corresponds to the ages of women currently invited for breast screening in most of the programmes. The second age group corresponds to women aged between 40 and 49. DG is calculated for each breast thickness by using the c-factors for the corresponding average

The impact of the new factors c and s on DG values obtained through the patients sample was analysed by recalculating the dose values obtained in previous studies (NHSBSP, 2003). It was found that, for the largest breasts (thickest on compression), the use of the c-factor increases doses by approximately 30%. For the smallest breasts, the dose estimates are decreased by 11%. The overall effect is to increase the average doses by about 11% for

This work took place within the framework of the Centre of Development of Nuclear Technology in Brazil through the project "Avaliação da Qualidade e Requisitos de Proteção Radiológica em Mamografia Digital e Monitoramento Dos Serviços de Mamografia de

ACR. American College of Radiology. (1998). Illustrated Breast Imaging Reporting and Data System (BI-RADS®), 3rd ed. *American College of Radiology*, Reston, Virginia, USA ACR. American College of Radiology. (1999). Mammography Quality Control Manual,

Alvarenga, FL and Nogueira MS. (2008). Análise de Parâmetros e Controle da Qualidade de

American Association of Physicists in Medicine. (2006). Acceptance testing and quality control

Argo WP, Hintenlang K, and Hintenlang DE. (2004). A tissue-equivalent phantom series for mammography dosimetry. *JACMP*. vol 5, N°4: 112-119. ISSN : 1526-9914 Beckers, S.W. et al. (2003). Results of technical quality control in the Dutch breast cancer

Beckett J and Kotre C J. (2000). Dosimetric implications of age related glandular changes in screening mammography. *Phys Med Biol,* 45 (2000): 801–813. ISSN 1361-6560 Boone, JM. (2002). Normalized glandular dose (DgN) coefficients for arbitrary x-ray spectra in

Sistemas de Radiologia Computadorizada para Mamografia. *Centro de Desenvolvimento da Tecnologia Nuclear*, Belo Horizonte, 2008. Available from http://www.dominiopublico.gov.br/pesquisa/DetalheObraForm.do?select\_action=

of photostimulable storage phosphor imaging systems. *AAPM Report n° 93*, October

screening programme(2001-2002). (Nijmegen, The Netherlands). Proc. EFOMP

mammography: Computer-fit values of Monte Carlo derived data. *Med. Phys*.

craniocaudal views and by about 14% for mediolateral oblique views.

*American College of Radiology*, Reston, Virginia, USA

Minas Gerais - apoiado pela FAPEMIG e PCI-CNPq."

**5.1.1.3 Measurements on patients** 

composition of each age group.

**6. Acknowledgments** 

&co\_obra=125026

2006. ISSN:0271-7344

29(2002):869–875. ISSN: 0094-2405

Congress.

**7. References** 

#### **5.1.1.2 The European and the IAEA approach: measurements of phantoms**

Dose assessment in mammography initially (CEC, 1996) relied on the estimation of DG for a 4.5 cm thick standard breast model with 50% glandularity in a central region by using a 4.0 cm thickness PMMA phantom. DG was estimated using the cG from Dance (1990) given in Table 1. It was recommended that patient dosimetry be performed by recording the exposure data and compressed breast thickness of at least 50 patients. Ka,i and DG were calculated for each patient using the cG factors tabulated by Dance (1990) for the corresponding HVL value and compressed breast thickness. The main problem associated with this methodology was firstly due to the fact that the average compressed breast thickness of a typical population is 5.5 cm. Secondly, several works showed a breast composition for the standard breast different to that of 50% glandularity. In addition, modern X-ray systems select different spectra as a function of both breast thickness and composition. In order to take into account all these factors, Dance et al. (2000) modify the definition of cG according to the expression given in Table 1. In this expression, the g-factor (unchanged from that initially used) corresponds to a glandularity of 50% and is tabulated for different breast thicknesses and HVL. The c-factor corrects for any difference in breast composition from 50% glandularity and is tabulated for different HVL, breast thicknesses and breast glandularities. The factor s makes a correction for the use of an Xray spectrum other than that for a Mo/Mo target–filter combination. The value for s depends only on the anode/filter combination, except in the case of W/Al which depends also on the kVp (Dance, 2009).

The equivalence between a range of PMMA thickness (2 - 8 cm) and compressed breast thickness has also been determined (Dance, 2000). According to the resulting equivalences, Ka,i delivered for a 4.5 cm thick PMMA phantom is equivalent to that for a 5.3 cm thick breast with 29% glandularity. This result was deduced from a sample of women in the age range 50-64 (Young, 1998; Becket, 2000). The mean glandular dose for different PMMA thicknesses is estimated using the relationship:

$$\mathbf{D}\_{\mathbf{G},\text{PT}} = \mathbf{K}\_{\text{PT}} \cdot \mathbf{g}\_{\text{GBT}} \cdot \mathbf{c}\_{\text{GBT}} \,\text{s} \tag{3}$$

*K* PT is the air kerma at the entrance surface of a phantom of PT thickness and the *gCBT* and *cCBT* factors are the values tabulated for the equivalent compressed breast thickness.

The DG limits proposed by the European (CEC, 2006) and IAEA (2011) protocols are given in Table 2 for a range of PMMA thickness. These values have been derived from screen/film mammography, since the cost associated with the transition to digital mammography should not imply an increase in the doses.


Table 2. Acceptable and achievable limits for mean glandular dose (DG).

#### **5.1.1.3 Measurements on patients**

128 Imaging of the Breast – Technical Aspects and Clinical Implication

Dose assessment in mammography initially (CEC, 1996) relied on the estimation of DG for a 4.5 cm thick standard breast model with 50% glandularity in a central region by using a 4.0 cm thickness PMMA phantom. DG was estimated using the cG from Dance (1990) given in Table 1. It was recommended that patient dosimetry be performed by recording the exposure data and compressed breast thickness of at least 50 patients. Ka,i and DG were calculated for each patient using the cG factors tabulated by Dance (1990) for the corresponding HVL value and compressed breast thickness. The main problem associated with this methodology was firstly due to the fact that the average compressed breast thickness of a typical population is 5.5 cm. Secondly, several works showed a breast composition for the standard breast different to that of 50% glandularity. In addition, modern X-ray systems select different spectra as a function of both breast thickness and composition. In order to take into account all these factors, Dance et al. (2000) modify the definition of cG according to the expression given in Table 1. In this expression, the g-factor (unchanged from that initially used) corresponds to a glandularity of 50% and is tabulated for different breast thicknesses and HVL. The c-factor corrects for any difference in breast composition from 50% glandularity and is tabulated for different HVL, breast thicknesses and breast glandularities. The factor s makes a correction for the use of an Xray spectrum other than that for a Mo/Mo target–filter combination. The value for s depends only on the anode/filter combination, except in the case of W/Al which depends also on the

The equivalence between a range of PMMA thickness (2 - 8 cm) and compressed breast thickness has also been determined (Dance, 2000). According to the resulting equivalences, Ka,i delivered for a 4.5 cm thick PMMA phantom is equivalent to that for a 5.3 cm thick breast with 29% glandularity. This result was deduced from a sample of women in the age range 50-64 (Young, 1998; Becket, 2000). The mean glandular dose for different PMMA

 DG, PT = K PT gCBT cCBTs (3) *K* PT is the air kerma at the entrance surface of a phantom of PT thickness and the *gCBT* and

The DG limits proposed by the European (CEC, 2006) and IAEA (2011) protocols are given in Table 2 for a range of PMMA thickness. These values have been derived from screen/film mammography, since the cost associated with the transition to digital mammography

> 20 21 1.0 0.6 30 32 1.5 1.0 40 45 2.0 1.6 45 53 2.5 2.0 50 60 3.0 2.4 60 75 4.5 3.6 70 90 6.5 5.1

Acceptable level for DG to equivalent breast (mGy)

Achievable level for DG to equivalent breast (mGy)

*cCBT* factors are the values tabulated for the equivalent compressed breast thickness.

Thickness of equivalent breast (mm)

Table 2. Acceptable and achievable limits for mean glandular dose (DG).

**5.1.1.2 The European and the IAEA approach: measurements of phantoms** 

kVp (Dance, 2009).

thicknesses is estimated using the relationship:

should not imply an increase in the doses.

Thickness of PMMA (mm)

The method followed for patient dosimetry relies on the results obtained in two studies (Young et al 1998; Beckett et al., 2000) that have each independently estimated the breast composition of women attending for screening. As a result, the average breast composition as a function of breast thickness was established for two age groups. One age group (aged 50 to 64) corresponds to the ages of women currently invited for breast screening in most of the programmes. The second age group corresponds to women aged between 40 and 49. DG is calculated for each breast thickness by using the c-factors for the corresponding average composition of each age group.

The impact of the new factors c and s on DG values obtained through the patients sample was analysed by recalculating the dose values obtained in previous studies (NHSBSP, 2003). It was found that, for the largest breasts (thickest on compression), the use of the c-factor increases doses by approximately 30%. For the smallest breasts, the dose estimates are decreased by 11%. The overall effect is to increase the average doses by about 11% for craniocaudal views and by about 14% for mediolateral oblique views.

## **6. Acknowledgments**

This work took place within the framework of the Centre of Development of Nuclear Technology in Brazil through the project "Avaliação da Qualidade e Requisitos de Proteção Radiológica em Mamografia Digital e Monitoramento Dos Serviços de Mamografia de Minas Gerais - apoiado pela FAPEMIG e PCI-CNPq."
