| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University Hospital Vienna and Medical University of
Vienna, Austria, Waehringer Guertel 18-20/7F, A-1090 Vienna, Austria.
2 Department of Biomedical Engineering and Physics, University Hospital Vienna
and Medical University of Vienna, Vienna, Austria.
Received October 11, 2004;
accepted after revision December 22, 2004.
Address correspondence to G. Schueller
(gerd.schueller{at}meduniwien.ac.at).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. For both a wet laser printer and a paper printer connected to an FFDM system, image quality parameters were evaluated using a standardized printer test image (luminance density, dynamic range). The detectability of standardized objects on a phantom was also evaluated. Furthermore, 640 mammograms of 80 patients with different breast tissue composition patterns were imaged with both printers. Subjective image quality parameters (brightness, contrast, and detection of details of anatomic structures—that is, skin, subcutis, musculature, glandular tissue, and fat), the detectability of breast lesions (mass, calcifications), and the diagnostic performance according to the BI-RADS classification were evaluated.
RESULTS. Both the luminance density and the dynamic range were superior for the wet laser printer. More standardized objects were visible on the phantom imaged with the wet laser printer than with the paper printer (13/16 vs 11/16). Each subjective image quality parameter of the mammograms from the wet laser printer was rated superior to those of the paper printer. Significantly more breast lesions were detected on the wet laser printer images than on the paper printer images (masses, 13 vs 10; calcifications, 65 vs 48; p < 0.05). With the paper printer images, BI-RADS 4 and 5 categories were underestimated for 10 (43.5%) of 23 patients.
CONCLUSION. For FFDM, images obtained from a wet laser printer show superior objective and subjective image quality compared with a paper printer. As a consequence, the paper printer should not be used for FFDM.
Keywords: digital imaging • full-field digital mammography • mammography • printers
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hard copies for FFDM are routinely generated by wet laser printers [16, 17], which have also become the state-of-the-art equipment for other imaging techniques such as CT, MRI, and sonography. However, wet laser images require chemical development, which is expensive, contributes to environmental pollution, and needs careful handling to obtain constant image quality. As an alternative to the wet laser printer systems, the use of paper printers has been initiated for the documentation of radiologic examinations [18]. In contrast to the complex printing requirements for the wet laser output, the handling requirements for paper printers are reduced to the delivery of paper. With this method, the environmental, time, storage, and delivery problems could potentially be eliminated. Most important, with increasing expenditures in medicine and increasing pressure from managed care, paper printers could markedly contribute to cost-cutting strategies in medical health care.
The purpose of our study was to determine whether a paper printer can reach the high standard of a wet laser printer for FFDM images. The image quality of both devices was assessed using objective criteria through the evaluation of a standardized printer test image and an ACR phantom [19]. In addition, subjective criteria were examined in a comparison of a total of 640 mammograms.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
The film used for the wet laser printer was a bluebase film (Scopix Laserfilm LT 2, Agfa) on which a silver grain emulsion is used to produce an image. A commercially available copy paper (120 g/m2, Colortec, Xerox) was used for the paper printer.
Objective Comparison
This part of the study was designed to compare the image quality parameters
of a standardized printer test image (GE Printer Test Image, GE Healthcare
[20]) and the ACR phantom
[19].
From the workstation of the Senographe 2000D, the GE Printer Test Image was printed with the wet laser printer and with the paper printer. The accuracy of specific image elements available on the GE Printer Test Image (lines, curves, and shapes) was determined by visual inspection. The luminance values (LV) of 21 gray-level steps were measured with a luminance meter (LS-110, Minolta) at a recommended distance from the test image of 120 cm [20]. Furthermore, the dynamic range (ratio of the highest to the lowest LV) of both printer devices was evaluated.
Exposure of the ACR phantom was performed on the FFDM system in automatic mode, and the processed image was printed with the wet laser printer and with the paper printer. The maximum numbers of visible small objects—that is, fibers, specks, and masses—were determined.
The evaluation of wet laser printer images was performed on a viewing box with a luminance greater than 3,000 cd/m2 (Rotolux 320 DS, Schulte) in the center, with a difference between the center and the corners of the viewing area of less than 15% [16, 21]. The evaluation of paper printer images was performed using a desk light with a 100-W bulb (Philips Medical Systems), following the manufacturer's guidelines.
Patients
From January 2003 to February 2004, 2,112 patients underwent FFDM at our
institution. The main indications for FFDM included the early detection of
breast cancer in asymptomatic patients and the diagnostic workup of
symptomatic patients. For this study, 80 patients with different breast-tissue
densities (BI-RADS composition pattern types) were randomly selected by a
computer search of our database. Thus, the final study population consisted of
20 patients of each BI-RADS composition pattern
[22]. The patients' ages
ranged from 20 to 78 years (mean age, 47 years).
Subjective Comparison
Mammography was performed with craniocaudal and mediolateral oblique
(30°) views of both breasts in all 80 patients. In total, 640 images were
obtained and then printed with the wet laser and paper printers. The
interpretation of the mammograms was performed under the same standardized
conditions as described for the objective comparison. Each of the images was
evaluated side by side by two experienced radiologists in an unlabeled, random
order to enable an accurate and direct comparison. In cases of discrepancy,
conclusions were reached by interrater consensus in the presence of a third
experienced radiologist. The reviewers were asked to compare the images on the
basis of the following aspects: brightness; contrast; and detail of anatomic
structures—that is, skin, subcutis, musculature, glandular tissue, and
fat. Each of the comparative aspects was categorized using a three-level scale
(good, moderate, or poor, with 1 being good). In case of breast lesions,
lesion characteristics—that is, mass or calcifications—were
evaluated. Furthermore, the diagnostic performance was assessed according to
the BI-RADS classification
[22].
Statistical Analysis
Statistical analysis was performed using commercially available software
(SPSS for Windows [Microsoft], release 11, Statistical Package for the Social
Sciences). For the objective comparisons, differences between the wet laser
printer and the paper printer in the visibility of small objects of the ACR
phantom were measured as sensitivity—that is, the ratio of visualized
and total lesions [23]. For
the subjective comparisons, for each aspect, agreement between the reviewers
was evaluated by means of the kappa statistic (Cohen's kappa test). Agreement
was defined as almost perfect, 
> 0.8; good, = 0.8-0.61;
moderate, = 0.60-0.41; fair, = 0.4-0.21; or poor, <
0.20 [23]. For the subjective
comparisons, the significance of differences between the two printers was
determined using the Wilcoxon's signed rank test. For all tests, the
significance level was set at p < 0.05
[23].
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ACR phantom—Results of visual inspection are summarized in Table 2. Five of six fibers were visible on both the wet laser printer image and the paper printer image. Of five specks, four were visible on the wet laser printer image and two on the paper printer image. On both images, four of five masses were visible. In total, with the wet laser printer, 13 of 16 small objects were visible, compared with 11 of 16 with the paper printer. The sensitivities of the wet laser printer and the paper printer are summarized in Table 2. The wet laser printer was more sensitive in the evaluation of specks (80% vs 40%) than the paper printer. No differences were measured in the evaluation of fibers (83%) and masses (80%). Consequently, the wet laser printer was more sensitive in the evaluation of all lesions than the paper printer (81% vs 69%). The differences were not significant (p > 0.05).
|
Subjective Comparison
Of a total of 2,240 subjective image quality scores per printer system (80
patients, 4 mammograms per patient, 7 scores), 2,228 scores (99.5%) rated a
good and 12 scores (0.5%) rated a moderate image quality for the wet laser
printer images. For the paper printer images, 248 scores (11.1%) rated a good,
584 scores (26.1%) rated a moderate, and 1,408 scores (62.9%) rated a poor
image quality. Only the image quality of the cutis was rated with a mean of
1.5 for the paper printer and, therefore, was not significantly different from
the results of the wet laser printer (Table
3). All other quality parameters of the wet laser printer were
significantly higher than those of the paper printer. The interrater
variability was almost perfect in the evaluation of each detail with the wet
laser printer; and almost perfect in two, good in one, and moderate in four
details with the paper printer (Table
3).
|
A total of 13 masses were seen with the wet laser printer (circumscribed, n = 3; indistinct, n = 2; spiculated, n = 6; others, n = 2) compared with 10 masses seen with the paper printer (circumscribed, n = 2; indistinct, n = 1; spiculated, n = 5; other, n = 2; p > 0.05). A total of 65 areas of calcifications were seen with the wet laser printer (amorphous, n = 17; pleomorphous, n = 11; linear, n = 5; other, n = 32) compared with 48 calcifications with the paper printer (amorphous, n = 12; pleomorphous, n = 3; linear, n = 3; others, n = 30; p < 0.05) (Table 4). Thus, three (23%) of 13 masses and 17 (26%) of 65 calcifications were not seen on paper printer images. When the images were classified using the BI-RADS categories, with the wet laser printer, BI-RADS category 1 was given in 19 cases, BI-RADS 2 in 31 cases, BI-RADS 3 in seven cases, BI-RADS 4 in 14 cases, and BI-RADS 5 in nine cases, compared with 23, 38, six, seven, and six, respectively, with the paper printer (Table 4, Figs. 2A, 2B, 2C, 2D, 2E, 2F, 2G, and 2H). Thus, with the paper printer, BI-RADS categories 4 and 5 were underestimated for 10 (43.5%) of 23 patients compared with the wet laser printer. The interrater variability was almost perfect in the evaluation of six and good in two of these details with the wet laser printer; and almost perfect in four and good in four details with the paper printer (Table 4).
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In FFDM, laser imagers have become the state-of-the-art equipment for printing hard copies of original data. They have undergone continuous improvement to offer better resolution, stability, and processing speed. Concomitantly, laser films have been improving in terms of dynamic range, archiving properties, and handling [17].
At first sight, several advantages of paper printer systems over wet laser systems for FFDM could be considered. First, a paper printer costs about 15% less than the wet laser printer. Second, the costs for produced films are about 10 times less with the paper printer system than with the wet laser printer system. In addition, with the wet laser printer, about 10% of costs per film sheet must be added for the development chemicals and for disposal of chemical waste material. Third, the installation and maintenance of the paper printer system are simple. Fluid or solid chemical waste is avoided, and therefore so are disposal problems, which reduces installation and operation costs. Finally, the paper printer images can be recycled, making this method environmentally friendly. The integration of inexpensive paper printer images to radiologic algorithms was described by Ibbott et al. [18]. They evaluated the image quality of paper printers and of wet laser printers, but a direct comparison of their results with ours would not be practical because only data from CT scans were included in their study.
However, for brightness and contrast, worse subjective impression on paper printer images correlated with measurable values in the objective comparison. This can be explained by the influence of the dynamic range on the reviewer's impression of contrast [24]. In our study, the dynamic range of the wet laser printer was superior to that of the paper printer. Thus, the gradation of the wet laser printer is significantly higher, and therefore the visibility of low-contrast elements is better, than with the paper printer. This is consistent with the results of the ACR phantom evaluation. The visibility of fibers and masses is equally high for both devices, but the visibility of specks (which emulate low-contrast calcifications) is better for the wet laser printer. Because of the low number of small objects on the ACR phantom, these differences are not significant. However, a trend toward a higher sensitivity with the wet laser printer than with the paper printer is observed. Consequently, when the visibility of small low-contrast elements is of real importance in FFDM, the paper printer is not suitable as a documentation tool.
The method used for imaging defines the performance of a system and the information that is available to the reviewer. In an outpatient setting especially, patients with abnormal findings are provided hard copies of their digital mammograms to obtain further explanation by the referring specialist, if necessary. Even if written reports were attached to these hard copies, we think that subtle, but nevertheless crucially important, details—for example, calcifications—would not be reproducible if the hard copies were printed by a paper printer. In our study, three (23%) of 13 masses and 17 (26%) of 65 calcifications were not detectable on paper printer images. As a consequence, with the paper printer, BI-RADS categories 4 and 5 ratings were underestimated for 10 (43.5%) of 23 patients compared with the wet laser printer. Thus, diagnostic or therapeutic procedures may be delayed in a considerable number of breast cancer patients.
The interrater agreement for the BI-RADS classification was excellent for
both printer devices ( = 0.9). Recently, Berg et al.
[25] reported an interrater
agreement for BI-RADS classification to be as low as 0.41-0.44, even after a
short period of BI-RADS training
[25]. The authors evaluated
the performance of 23 practicing mammogram-interpreting physicians. In
contrast, in our study the mammograms were examined by two experienced
radiologists who have used the BI-RADS system for several years. This may
explain the differences in the interrater agreement between our study and the
results reported in the literature.
As a limitation of our study, the superior detectability of small objects with the wet laser printer, particularly in low-contrast tissue, may be relevant in the following respect: Because we selected patients using a computer search of our database, there were equal numbers of patients with different breast-tissue composition patterns. However, in a routine clinical setting, one would encounter more mammograms with extremely-low and low breast-tissue density, and consequently find slightly better results for the paper printer. However, we think that our data are reliably unequivocal enough to draw conclusions for hard-copy printing in FFDM.
In summary, images obtained from a wet laser printer show superior objective and subjective image quality compared with a paper printer. The advantages of the paper printer, particularly in terms of cost reduction, do not compensate for low image quality. As a consequence, the tested paper printer should not be recommended for FFDM.
Acknowledgments
We thank Mary McAllister, John Hopkins University Hospital, Baltimore, MD,
for manuscript assistance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. A. KHALIFAH, A. BRINDHABAN, G. A. BALOUL, F. A. BATHER, and M. ABDULGAFOOR Optical Density Changes In Dry-processed Films Radiol. Technol., September 1, 2007; 79(1): 9 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. M. Hall Wet Versus Dry Laser Printers for Copying Digital Mammograms Am. J. Roentgenol., June 1, 2006; 186(6): E22 - E22. [Full Text] [PDF]
|
|
|
|
|
|
G. Schueller Reply Am. J. Roentgenol., June 1, 2006; 186(6): E22 - E22. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
