| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Full paper |
1National Co-ordinating Centre for the Physics of Mammography and 2Regional Radiation Protection Service, Department of Medical Physics, St Luke's Wing, Royal Surrey County Hospital, Guildford GU2 7XX, 3Warwickshire, Solihull & Coventry Breast Screening Centre, Coventry and Warwickshire Hospital, Stoney Stanton Road, Coventry CV1 4FH and 4Jarvis Breast Screening Centre, 60 Stoughton Road, Guildford GU1 1LJ, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Fuji (Fuji Photo Film (U.K.), London, UK) and Kodak (Kodak Ltd., Hemel Hempstead, UK) have developed high contrast film–screen combinations for mammography. The Fuji AD-M imaging system comprises the AD-M film together with either fine or medium AD-M screens. Fuji AD-M film has two emulsion layers on a single side of film. These are designed to provide a steep characteristic curve down to a relatively low optical density (OD). The aim of this design is to improve the visualization of dense glandular tissues on mammograms. The Fuji AD-M imaging system was compared with the Kodak imaging system, which uses min-R 2000 film with min-R 2000 (fine) or min-R 2190 (medium) screens. The objective of this study was to evaluate the implications of using these high contrast film–screen combinations in the NHSBSP. Fuji UM-MA HC film combined with fine screens was used for comparison since it has a lower contrast than both Fuji AD-M and Kodak min-R 2000, and is still in widespread use in the NHSBSP [1].
Fuji AD-M and Kodak min-R 2000 films have been designed to produce high contrast mammograms. Whilst high contrast has advantages, it can also lead to problems owing to reduced latitude. A quantitative method to investigate potential latitude problems was used in earlier work [2]. Previously, the dynamic range (DR) in the main breast region was calculated for each mammogram as the difference between the log10 relative exposures corresponding to the maximum and minimum ODs. The DR was also used to measure the latitude available with a specific film–screen combination, and to estimate what percentage of a sample of mammograms recorded using that combination would be expected to be imaged optimally. Very high film contrast is not expected to be a problem when visualizing fatty breasts, but this may not be the case when trying to visualize areas of dense glandular tissues. It is conceivable that it may be more difficult to locate suspicious masses in some dense tissues (shown at low OD) when very high contrast films are used. A new measure (Llower) is used here to quantify the latitude available from mid to low OD with the three film–screen combinations considered.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Comparison of film–screen combinations
A contralateral study was performed using mammograms from 114 symptomatic women referred for mammography. For each woman, two views were recorded for each breast using routine patient exposures for diagnostic purposes. For each view, one breast was imaged using the Fuji UM-MA HC film–screen combination and the other using one of four Fuji AD-M combinations of screen and processing cycle. The film–screen combinations used are described in Table 1
. 452 mammograms (235 medio-lateral oblique views and 217 cranio-caudal views) were studied. The women's ages ranged from 30 years to 71 years, with a mean (±SD) of 51.6 years±7.0 years.
|
A technical evaluation of the film–screen combinations was also undertaken. To compare the characteristic curves, a calibrated aluminium step-wedge was imaged at 28 kV, using a Mo/Mo target/filter combination, for each of the film–screen combinations used. The relative speeds of each film–screen combination were obtained by comparing the exposures required to produce a film OD of 1.6 for a 4 cm thickness of PMMA.
Visual grading analysis
The mammograms were viewed in low ambient light conditions on a masked light box normally used for reviewing films, by an expert radiologist with considerable experience in mammography. For each mammogram, breast composition was graded as either fatty, mixed density or dense. The radiologist's opinion of the OD of glandular tissue, adipose tissue, pectoral muscle, and the skin edge were graded on a seven point scale from very high to very low with the mid-point, OK representing the ideal. An assessment was made of the relative image quality of the pairs of mammograms to determine whether the mammogram using the Fuji AD-M film was better, the same or worse than the mammogram using theFuji UM-MA film, and whether or not the mammograms were acceptable. The overall diagnostic value of each mammogram covering film exposure, contrast, OD, noise and artefacts, was graded as either excellent, good, satisfactory, poor or inadequate.
Evaluation of films in a screening environment
For the second part of the clinical evaluation, samples of Fuji AD-M and Kodak min-R 2000 mammograms were collected from two NHSBSP centres for analysis.
Fuji AD-M imaging system
116 mammograms (84 mediolateral oblique views and 32 craniocaudal views) were collected from 42 women attending a single day of screening on a mobile unit, using the Fuji AD-M film with Fuji AD-M fine screens. The women's ages ranged from 46 years to 63 years, with a mean age (±SD) of 54.3 years±4.3 years.
Mammograms were taken using a Siemens Mammomat 2S (Siemens plc Medical Engineering, Bracknell, UK), operated under AEC. A tube potential of 28 kV was employed for all exposures, and a Mo/Mo target/filter combination was used together with the bucky grid. A target OD of 1.6 for a 4 cm thickness of PMMA was used and the correct operation of the AEC checked. The AEC chamber was placed at the chest wall position for all films. The films were processed using a Fuji 2800 processor operating at a developer temperature of 34 °C on an extended cycle using Photosol chemicals. The films were batch processed at the end of the day's screening.
To compare the characteristic curve of the Fuji AD-M film with the characteristic curves of the other film types used in this study, a calibrated aluminium step-wedge was imaged at 28 kV (using a Mo/Mo target/filter combination) on the same day as screening.
Kodak min-R 2000 imaging system
122 mammograms (89 mediolateral oblique views and 33 craniocaudal views) were collected from 45 women attending a single day of screening on a mobile unit, using Kodak min-R 2000 film with Kodak min-R 2000 screens. The women's ages ranged from 49 years to 64 years, with a mean age (±SD) of 55.7 years±4.3 years.
Mammograms were taken using a GE Medical Systems Senographe 600 TS operated under AEC. A tube potential of 28 kV was used for the majority of exposures. However, 30 kV was selected for a small number of women with large breasts, at the discretion of the radiographers. The 30 kV films were analysed in the same way as the 28 kV films, with the exception of the characteristic curve data. A Mo/Mo target/filter combination was used together with the bucky grid. A target OD of 1.6 for a 4 cm thickness of PMMA was used and the correct operation of the AEC checked. The AEC chamber was positioned at the chest wall for the quality control films and moved, depending on breast size, for the patient films. Films were processed using a Kodak M35M processor operating at a developer temperature of 35 °C on a standard cycle (150 s dry to dry) and using Kodak RP X-Omat EX chemicals. The films were batch processed at the end of the day's screening.
To compare the characteristic curve of the Kodak min-R 2000 film with the characteristic curves of the other film types used in this study, a calibrated Aluminium step-wedge was imaged at 28 kV (using a Mo/Mo target/filter combination) on the same day as the screening.
Quantitative assessment of image quality in mammograms
Both sets of mammograms were quantitatively assessed using an existing protocol, which has been described in detail elsewhere [2]. The mammograms were digitized using a Lumiscan 150 HR laser scanner (Lumisys, Sunnyvale, CA) at 210 µm resolution. The scanner's detected signal was digitized into image pixel values. Grey scale step-wedge films were digitized with the patient films to enable conversion from image pixel value to OD units. The image processing software package Aphelion (Amerinex Applied Imaging, Inc., Northampton, MA) was used to create and store a pixel value to OD calibration. This was then used to produce images with pixels in OD units.
Digitized images of mammograms were analysed using mutually exclusive regions of interest (ROI) representing the pectoral muscle (oblique films only), main breast and skin edge. A semi-automatic means of selecting the ROIs was again used [2]. However, before the threshold function was used to define the ROIs, the images were reduced in size by a factor of two and reduced to 8 bits to increase processing speed and Gaussian filtered (Gaussian width 11) to smooth the image. The ROIs selected by the threshold technique were resized to fit the original images.
Image analysis performed on each digitized mammogram included measuring the maximum, minimum and mean OD in each ROI. The maximum and minimum ODs were measured after the images had been filtered twice using an octagonal (11 x 11, 97 element) median filter to remove random and "shot" noise. Shot noise is present in an image when individual pixels are corrupted or missing from the image, and the octagonal median filter replaces these pixels with a value from the median neighbourhood [3]. The filter was used to remove calcifications, film emulsion pick-off, random noise and artefacts caused by foreign bodies, e.g. dust, on the screens. A contrast index was calculated for each mammogram as the difference between the maximum and minimum ODs in the main breast ROI. The log10(relative exposure) (log10E) required to produce each OD value was calculated using the measured characteristic curves of the film–screen combinations [2].
The shapes of the low OD "toe" end of the characteristic curves of the film–screen systems used in this study differ and therefore their ability to adequately show dense glandular tissues may also differ. To quantify these differences a measure of the latitude available in the lower half of the curve, Llower, was used. Llower was calculated for each film–screen system using
|
|
where Emean is the relative exposure corresponding to a mean OD of 1.8 and Emin is the relative exposure corresponding to the suggested minimum OD limit for each film type. For all three film–screen combinations an OD of 1.8 was used for Emean, as it appeared to be close to the typical mean OD used in practice in the UK NHSBSP [4, 5]. It was also judged by the radiologists to be the ideal mean OD. Emean corresponds to a mean OD in the main breast region of a mammogram. Mean ODs in the main breast region were higher than the film ODs for QA films in this and previous studies [2, 5]. The difference between the mean OD in the main breast region and the OD measured on QA films was approximately 0.2 OD. The Emin limits were determined from minimum OD measurements in mammograms and the results of visual grading analysis performed by radiologists. The lower dynamic range (DRl) of each mammogram was calculated using
|
|
The percentage of mammograms that have a "lower" dynamic range greater than Llower, and may not therefore be imaged optimally, was then calculated for each film–screen combination.
To study skin edge densities, two sets of OD measurements were made using a calibrated X-rite 301 spot densitometer (X-rite, Congleton, UK), with a 2 mm diameter circular spot. The first set of measurements were taken to determine the maximum OD in the skin edge ROI, and a second set of measurements were taken to determine the maximum OD in the fully exposed black part of the mammogram, as described by Meeson et al [6].
Visual grading analysis
Two expert radiologists, with considerable experience in mammography, independently reviewed each mammogram. In addition to grading breast composition and OD, as previously described, the mammogram assessments included grading the skin edge region as visible, just visible, or too dark [6]. The relationships between quantitative measures of image quality (OD and contrast) and the opinions of the radiologists were then examined.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
. The average gradient between an OD of 1.0 and 2.0 above base plus fog was 3.67 for Fuji UM-MA HC film and 4.61 for Fuji AD-M film (using fine screens and extended cycle processing). The gradient between an OD of 0.25 and 2.00 above base plus fog was 3.10 for Fuji UM-MA HC film, and 3.89 for Fuji AD-M film. Thus Fuji AD-M film had a contrast that was about 25% higher than Fuji UM-MA HC film. In Table 2 the differences in gradient between the two Fuji AD-M film–screen combinations were small, and the rapid cycle processing produced slightly higher contrast than the extended cycle processing. The relative speed data in Table 2 have been normalized to a value of 100 for the Fuji UM-MA HC film–screen combination. The relative speeds of Fuji UM-MA HC and Fuji AD-M combination 1 (fine screen and extended cycle processing) were comparable. Comparing Fuji AD-M film combinations with the same processing cycles, the relative speeds with the fine screens are approximately 30% less than those measured with the medium screens.
|
. For all curves, a log10E of zero corresponds to that required for an OD of 1.6. The Fuji AD-M imaging system and Kodak min-R 2000 imaging system curves had similar gradients, excluding the average gradient between an OD of 0.25 and 2.00 above base, which was less for the Kodak min-R 2000 imaging system (average gradient of 3.77) than for the Fuji AD-M imaging system. It can be inferred from Figure 1 that both Fuji AD-M and Kodak min-R 2000 films have substantially higher contrast than Fuji UM-MA HC film. The gradients of the characteristic curves for each film are plotted against OD in Figure 2. Kodak min-R 2000 and Fuji AD-M films have higher gradients than Fuji UM-MA HC film across a wide range of OD (OD
1.0). Below an OD of 1.0, Kodak min-R 2000 film and Fuji UM-MA HC film are similar while Fuji AD-M film has a notably higher gradient. Figure 3 illustrates the relationship between exposure and film gradient. The difference in gradients between Fuji UM-MA HC film and the other two films is again apparent across the range of log10E corresponding to the mid-range of ODs (between a log10E of –0.2 and 0.2). The gradient of Fuji AD-M film crosses over from above to below that of the other two films for exposures between approximately –0.2 and –0.3 (log10E).
|
|
|
. The majority of mammograms have a DRl between approximately 0.1 and 0.3. The Llower was 0.28 for Fuji UM-MA HC film, 0.26 for Fuji AD-M film and 0.22 for Kodak min-R 2000 film. DRl was greater than the Llower in 18% of mammograms for Fuji UM-MA HC film, compared with 26% of mammograms for Fuji AD-M film and 44% of mammograms for Kodak min-R 2000 film.
|
The most striking difference between mammograms generated using Fuji AD-M film and Fuji UM-MA HC film was usually the apparently sharper appearance achieved with Fuji AD-M film. This apparent increase in visual sharpness was thought mainly to be owing to higher film contrast. Figure 5 shows how the overall diagnostic value of the mammograms varied with the film type used. The majority of the mammograms produced with Fuji UM-MA HC film were graded as satisfactory, whereas the majority of mammograms produced with Fuji AD-M film were judged to be in the higher good category. While the mammograms generated using Fuji UM-MA HC film were either satisfactory or good, only Fuji AD-M film produced any poor or inadequate mammograms. Overall, 90.5% of Fuji AD-M mammograms were judged to be good, but unlike Fuji UM-MA HC film, a few Fuji AD-M mammograms (2.3%) were deemed to be unacceptable for mammography owing to unfavourable visualization of dense glandular tissues.
|
shows the relationship between breast composition and whether Fuji AD-M mammograms appeared better overall than Fuji UM-MA HC mammograms. The data for each breast type have been normalized and plotted as percentages. Fatty breasts were all visualized better with Fuji AD-M film. However, 16% of dense breasts were visualized better with Fuji UM-MA HC film rather than with Fuji AD-M film, with a small number of mixed density breasts being visualized worse with Fuji AD-M film compared with Fuji UM-MA HC film.
|
Fuji AD-M film was always judged to be superior to Fuji UM-MA HC film for visualizing breasts which were predominantly fatty. However, pectoral muscle visualization was better with Fuji UM-MA HC film than with Fuji AD-M film. For both film types the majority of the mammograms had skin edge regions that were judged too dark (98% of Fuji AD-M films and 91% of Fuji UM-MA HC films).
Evaluation of films in a screening environment
Film OD
Table 3 contains the range of mean OD and the average mean OD in the main breast ROIs for mammograms produced using Fuji AD-M and Kodak min-R 2000 films. The ranges of mean OD were similar for both film types. The average mean OD for the Fuji AD-M imaging system was slightly lower than for the Kodak min-R 2000 imaging system. However, both systems complied with the NHSBSP requirement that standard film densities measured with PMMA blocks be maintained within ±0.20 of their target density, which was 1.6 in both cases [4].
|
for each grade of skin edge visualization. For both imaging systems the skin edge ODs were close to the ODs in the black part of the mammogram when the skin edge was dark and the skin line could not be seen. Viewed under normal viewing conditions the skin edge regions were visible up to an OD of approximately 3.6 for both film types.
|
for both Fuji AD-M and Kodak min-R 2000 imaging systems. The dense breasts had the greatest mean contrast index for both film systems. For mixed and fatty breasts the mean contrast indexes were slightly greater for the Fuji AD-M imaging system than for the Kodak min-R 2000 imaging system.
|
for both Fuji AD-M and Kodak min-R 2000 films. For each grade of glandular tissue OD, Fuji AD-M film had lower minimum ODs than Kodak min-R 2000 film. 95% of mammograms graded as having areas of low glandular tissue OD had a minimum OD below approximately 0.8 for Kodak min-R 2000 film and below approximately 0.6 for Fuji AD-M film. The mean minimum ODs for each grade of glandular tissue visualization are listed in Table 6 for both imaging systems. The mammograms where the ODs of glandular tissue were graded as low or slightly low had a mean minimum OD below approximately 0.8, and are mammograms from the dense breast group or the mixed density breast group with local dense glandular tissues. The mean minimum ODs for Fuji AD-M film were of the order of 0.2 OD lower than for Kodak min-R 2000 film. In a similar test to assess the acceptable maximum density in the main breast region, tissue density was generally graded as slightly high or high above an OD of 2.9 for Kodak min-R 2000 film, but there was no clear relationship for Fuji AD-M film. The mean minimum ODs in the main breast region for each breast type are shown in Table 7. For both film types there were far more mixed density breasts than dense or fatty breasts. The difference in mean minimum OD between the two film types for the mixed density breasts was of the order of 0.2 OD. The difference for the other two breast types was less marked. The mean minimum OD for each breast type, subdivided by the radiologists' grading of the OD of glandular tissue, is shown in Table 8. The mean minimum OD of the dense breasts imaged with Fuji AD-M film decreased as the radiologists' opinion of glandular tissue OD decreased from OK to low. For the mixed density breasts the larger samples of data make the inter-film type comparison more significant. The ODs of glandular tissue were judged to be OK at OD greater than approximately 0.6 and 0.8 for Fuji AD-M film and Kodak min-R 2000 film, respectively. For minimum ODs below these values the visualization of structure in glandular tissue can become more difficult. For all breast types, the grades of the OD of glandular tissue have lower mean minimum ODs with Fuji AD-M film than with Kodak min-R 2000 film. With few exceptions fatty breasts have mean minimum ODs that are well above the corresponding suggested minimum OD limits.
|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
The average film gradients between an OD of 1.0 and 2.0 indicate that while Fuji UM-MA HC film is typical of the films used in the NHSBSP until recently [7], Fuji AD-M film is a much higher contrast film. The higher contrast of Fuji AD-M film can be seen in Figure 1
, and when quantified, Fuji AD-M film had approximately 25% more contrast than Fuji UM-MA HC film.
Both Fuji AD-M and Kodak min-R 2000 are very high contrast films. Fuji AD-M film generally had the highest film gradient over the mid-range of exposures, and therefore higher contrast was achieved with Fuji AD-M film. However, as shown in Figure 3 the Fuji AD-M film gradient fell very rapidly at lower exposures corresponding to an OD of less than approximately 0.5 (equivalent to a log10E of –0.3). At these exposures the film contrast was less than with the other film types. The suggested minimum OD limits for ideal glandular tissue visualization are 0.8 for Kodak min-R 2000 film and 0.6 for Fuji AD-M film. The limits are based on radiologists' opinions, but appear to be related to the lower film contrast below these limits. (The minimum OD limits correspond to gradients of approximately 2.5 to 2.8.) In another study a minimum OD limit of 0.8 was suggested for Fuji UM-MA HC film [5]. Below these suggested OD limits it becomes increasingly difficult to visualize structure within dense glandular tissue.
Llower is a measure of the latitude available from mid to low OD. Both Fuji AD-M and Kodak min-R 2000 films have smaller Llower than Fuji UM-MA HC film. However, the Llower of Fuji AD-M film is closer to the Llower of Fuji UM-MA HC film than to the Llower of Kodak min-R 2000 film. This is a consequence of the design of Fuji AD-M film. The unusual shape of the low OD "toe" end of the Fuji AD-M film's characteristic curve effectively increased the available latitude. The percentage of mammograms with a DRl greater than the corresponding Llower was smallest for Fuji UM-MA HC film, the film type with the lowest contrast. This film type will therefore have the fewest mammograms that cannot be optimally imaged at any exposure. Screening centres using both Fuji AD-M and Kodak min-R 2000 films are more likely to have mammograms in which dense glandular tissue may be judged to be too light.
Comparison of film–screen combinations
Fuji AD-M film was designed to have very high contrast and improved breast tissue visualization at low optical density. To determine if these changes can be observed clinically, Fuji AD-M film was radiographically compared with Fuji UM-MA HC film in a contralateral study of mammograms from a group of women who had been referred for mammography. The women in the study were symptomatic patients with an age distribution not typical of the UK NHSBSP. The sample includes women below the minimum age for the NHSBSP. The likely effect of this is more women with dense glandular breast tissues being included in the study than would otherwise be expected. However, since a contralateral study was performed to compare the two Fuji film types, the results remain valid in indicating the potential benefits and problems of using Fuji AD-M film instead of Fuji UM-MA HC film.
Radiological assessment found that although all fatty breasts were visualized better with Fuji AD-M film, 16% of dense breasts were visualized worse with Fuji AD-M film. The Fuji AD-M films that appeared worse than the Fuji UM-MA HC films had contrast that was judged to be slightly high to very high, and the ODs of glandular tissue were judged to be slightly low to very low. Such effects are as expected owing to the differences in film latitude for the film types.
Overall, comparing Fuji AD-M film with Fuji UM-MA HC film, Fuji AD-M film generally appeared better. However, for visualizing the pectoral muscle Fuji UM-MA HC film appeared better than Fuji AD-M film. Both films produced mostly dark skin edge regions. With both films there were many instances when the actual breast outline could not be visualized clearly without the use of a bright light, and the area of breast tissues visible in clinical viewing conditions appeared to be less with the Fuji AD-M film. However, this is a difficulty that may be tolerated by many radiologists.
In terms of overall diagnostic value, Fuji AD-M film was generally considered to be superior. However, only Fuji AD-M film produced poor or inadequate mammograms and these were due to excessive contrast. Overall, 2.3% of the Fuji AD-M mammograms were classed as unacceptable and might have led to a technical recall of the patient.
The occasional problems associated with visualizing some areas of dense glandular tissue with Fuji AD-M film mean that either a different film–screen combination should be used to produce mammograms in these instances or the imaging equipment and procedures to produce the mammogram should be modified to improve the imaging of dense tissues. Using previous knowledge of breast composition, it may be possible to identify which women would require a modified protocol based upon earlier mammograms. When high contrast films are used, the target density used in setting up the X-ray set's AEC system has a marked effect on the final quality of the films produced. This should be high enough to allow adequate visualization of dense glandular tissues. Furthermore, the correct operation of the AEC is essential if the required mean OD is to be achieved.
Evaluation of films in a screening environment
Samples of Fuji AD-M and Kodak min-R 2000 mammograms were digitized and analysed to produce indices of OD, contrast and exposure. Visual grading analysis was conducted independently on each mammogram by two radiologists. These subjective judgements, together with quantitative measurements, have been utilized to assess the implications of using these films in the NHSBSP and whether the modified design of the Fuji AD-M film improves the visualization of breast tissues at low OD.
Two of the women screened at the centre using Fuji AD-M film were younger, aged 46 years, than the women who are routinely screened in the UK NHSBSP. These two women were screened as part of a national age trial. While these women may have had more dense glandular breast tissues than would otherwise have been expected, their inclusion in a study to determine whether areas of dense glandular tissues are adequately visualized should not cause bias.
The difference in mean minimum OD between the Fuji AD-M and Kodak min-R 2000 film types was of the order of 0.2 OD (for mixed density breasts), and the difference between the suggested lower OD limits was also 0.2 OD. The overall mean ODs for these two film types are slightly different, as shown in Table 3. This difference, however, does not solely account for the 0.2 OD difference in minimum ODs and limits. The shape of the "toe" of the characteristic curve for Fuji AD-M film was probably the main factor.
The curves in Figure 1 allow the effective dynamic ranges of the film types to be estimated using the recommended OD limits and the ideal mean film OD. The suggested lower OD limits are based on the mammograms judged by the radiologists to have low glandular tissue OD. For each film type 95% of the mammograms graded as having areas of low glandular tissue OD have minimum ODs below the suggested lower OD limits. However, 51% of the Fuji AD-M films had minimum ODs below an OD of 0.6, as compared with 10% of the Kodak min-R 2000 films. These percentages include some mammograms in which the ODs of glandular tissues werejudged to be OK by the radiologists. Consequently, visualizing glandular tissues satisfactorilyis not only dependent on the minimum OD in the main breast region, other factors must also be involved, e.g. the area of the mammogram with a particular OD.
The ideal range of OD for optimized breast tissue visualization in the main breast was judged to be 0.8 to 2.9 in an earlier study using Sterling Microvision C film (Sterling UK, Stevenage Herts., UK) [2]. These limits also appear suitable for Kodak min-R 2000 film and Fuji UM-MA HC film. However, glandular tissues could be adequately displayed down to lower OD with the Fuji AD-M film. Work reported from the Netherlands Breast Screening Programme also suggested that, in general, the minimum OD of glandular tissue should be no less than 0.8 [8]. A case can now be made for reviewing the European Criteria [9], which state that the useful range of ODs in mammography is 0.5–2.2, with a view to modifying the limits [5] where appropriate.
It has previously been found that for good skin edge visualization the maximum OD in the skin edge region should be less than 3.8 and there should be a large difference between the maximum OD in the skin edge region and the maximum OD of the fully exposed black part of the mammogram [6]. These results are supported by the findings in this study. Skin edge regions were generally visible for ODs of approximately 3.6 or less. Where the skin edge regions were judged to be dark and the skin line could not be seen, the OD was greater than 3.6 and the maximum OD in the skin edge region was close to the maximum OD of the fully exposed black part of the mammograms for both Fuji AD-M and Kodak min-R 2000 films.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
|
Acknowledgments |
|---|
Received for publication June 15, 2000. Revision received December 15, 2000. Accepted for publication January 31, 2001.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A D Dimakopoulou, I A Tsalafoutas, E K Georgiou, and E N Yakoumakis Image quality and breast dose of 24 screen-film combinations for mammography Br. J. Radiol., February 1, 2006; 79(938): 123 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I A Tsalafoutas, A D Dimakopoulou, E D Koulentianos, A N Serefoglou, and E N Yakoumakis Variation of the sensitometric characteristics of seven mammographic films with processing conditions Br. J. Radiol., August 1, 2004; 77(920): 666 - 671. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S Meeson, K C Young, M G Wallis, J Cooke, A Cummin, and M L Ramsdale Image features of true positive and false negative cancers in screening mammograms Br. J. Radiol., January 1, 2003; 76(901): 13 - 21. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
