Digital mammography: current view and future applications

Abstracts

Na mamografia digital, os processos de aquisição da imagem, demonstração e armazenamento são separados, o que leva à otimização de cada uma dessas etapas. A radiação transmitida através da mama é absorvida por um detector eletrônico, em resposta fiel a uma ampla variedade de intensidades. Uma vez que esta informação é armazenada, ela pode ser demonstrada usando técnicas computadorizadas de imagem, permitindo variações de brilho e contraste e ampliação, sem a necessidade de exposições radiológicas adicionais para a paciente. Neste artigo, o estado atual da tecnologia em mamografia digital e dados sobre testes clínicos que dão suporte ao uso dessa tecnologia são revistos. Além disso, algumas aplicações potencialmente utilizáveis que estão sendo desenvolvidas com a mamografia digital são descritas.

Mamografia; Câncer de mama; Radiografia digital; Mamografia digital

UPDATING

Digital mammography: current view and future applications^*

Andréa Gonçalves de Freitas^I; Cláudio Kemp^II; Maria Helena Louveira^I; Sandra Maria Fujiwara^III; Leandro Ferracini Campos^III

^IMD, Radiologists, Doctors in Medicine by Department of Imaging Diagnosis, Sector of Breast Radiology, Universidade Federal de São Paulo-Escola Paulista de Medicina, Titular Members of Colégio Brasileiro de Radiologia e Diagnóstico por Imagem

^IIAdjunct Professor at Department of Gynecology, Chief Sector of Breast Radiology, Department of Imaging Diagnosis, Universidade Federal de São Paulo-Escola Paulista de Medicina

^IIIMD, Radiologists, Titular Members of Colégio Brasileiro de Radiologia e Diagnóstico por Imagem

^{Mailing Address}

ABSTRACT

In digital mammography, imaging acquisition, display and storage processes are separated and individually optimized. Radiation transmitted through the breast is absorbed by an electronic detector with an accurate response over a wide range of intensities. Once these data are stored, they can be displayed by means of computer image-processing techniques to allow arbitrary settings of image brightness, contrast and magnification, without the need for further radiological exposure of the patient. In this article, the current state of the art in technology for digital mammography and clinical trials data supporting the use of this technology are reviewed. In addition, several potentially useful applications, currently under development with digital mammography, are described.

Keywords: Mammography; Breast cancer; Digital radiography; Full-field digital mammography.

INTRODUCTION

In the last years, there has been an increasing concern about the improvement of the technology involving the quality of mammographic images, characterized especially by a better contrast of structures to be analyzed, since both normal and pathological breast tissue present a similar radiological density. Main factors possibly limiting such structures contrast include beam energy, intensifying screen-film combination, film processing, quantity of radiation measured in milliampere per second (mAs), visualization conditions, besides the fact that the film is simultaneously a image receptor, a visualization media and a long term storage medium. These limitations may lead to a loss of image contrast, especially when the film exposure or processing conditions result in a decrease of optical density in tissues containing lesion^(1,2).

Studies related to the analysis of the conventional mammographic performance show that one of the greatest problems involving loss of imaging quality, with consequent patients recall, occurs because of images processing failure such as chemicals contamination or problems related to cleaning of dark chambers and intensifier screens, since dust causes artifacts, affecting the diagnosis^(1,2).

One the recent mammography developments approved by the US Food and Drug Administration (FDA) in January 2000, is the full-field digital mammography, where a set of semiconductors is the detector, replacing the X-ray film and receiving the radiation, transforming it into an electrical signal that, on its turn, is transmitted to a computer⁽³⁾. The North American National Cancer Institute has designated the full-field digital mammography as the diagnostic imaging technology with highest potential to improve the breast cancer detection and diagnosis⁽⁴⁾.

The elimination of the film limitation and utilization of the post-acquisition image processing resource (after radiological exposure) will reduce considerably the number of unsatisfactory images, resulting in a reduction of the population radiological overexposure and, consequently, time and costs involved in repetition of technically unsatisfactory images.

Some studies on physical features of the full-field digital detector have shown favorable results regarding its spatial resolution and detective quantum efficiency (DQE)^(5,6); other studies demonstrate even better results of contrast details⁽⁷⁾.

The majority of current researches directed to digital mammography present comparative parameters between this method and the conventional mammography, because this system is a reference in detection of early-stage breast lesions. There is a number of prospects in regard to the progress of the digital mammographic imaging method in relation to the conventional mammography and especially to its benefits for patients.

TECHNICAL CONSIDERATIONS AND FUNDAMENTALS OF PHYSICS

The increase of digital mammography relevance has been parallel to the increase of interest in digital images of other organic systems. The digital technology was initially utilized in mammography-guided interventional procedures, using detectors with small field-of-view (FOV). This development was delayed because of the necessity of a higher spatial resolution for mammographic images than for other radiological studies, besides the necessity of an increase in the level of efficiency in detection, representation of X-ray photons and of a low-noise system^(4,8–11). The direct acquisition digital mammography was initially created with systems of chips based on the CLD (coupled load device) technology utilized in stereotaxis for preoperative localization and biopsies^(12,13). These chips, however, could not be utilized for full-field digital mammography due to the detector restricted size. The subsequent attempt to perform a digital mammography was that of utilizing digitalizing systems with storage phosphor plates, denominated CR (computed radiography) systems. This technique, however, has not been established because of the limited DQE and low spatial resolution⁽¹⁴⁾. So, aiming at increasing the spatial resolution, one has investigated a technical combination of direct magnification with a microfocus and storage phosphor plates⁽¹⁵⁾. Results have been encouraging and the mammographic magnification with this technique has demonstrated to be superior to the conventional mammographic magnification for microcalcifications detection^(9,16,17). A new development of this combination, however, occurred late in the nineties. Presently, high resolution digitized storage phosphor systems are available and preliminary studies with phantoms have demonstrated promising results. Other full-field digital systems are currently being developed to overcome the first digital devices limitations. These systems characteristics vary: the spatial resolution ranges between 41 µm and 100 µm/pixel and the images contrast resolution ranges between 10 and 14 bits/pixel.

Films utilized in conventional mammography have high spatial resolution, between 12 and 15 pairs of lines/mm, but this resolution is limited because of the decrease in the capacity to distinguish structures with little contrast, whose X-ray absorption coefficient hardly differ. Also, a greater limitation occurs when problems resulting from conventional mammography exposure and processing conditions are associated^(1,2). As a result, researches have been performed in an attempt to develop digital images receivers for replacing the screen-film system currently utilized in mammography^(18,19). For digital systems, the spatial resolution ranges from five to tem pair of lines/mm, but they have a better contrast resolution⁽²⁰⁾.

Amongst the physics prospects in relation to full-field digital mammography, there is an expectation of improvement in lesions characterization with reduction of the radiation dose received by the patient. Initially, this is due to the increase in the efficiency of X-ray photons detection in a low-noise system as a result o f a linear response to the incident radiation intensity, decrease in the number of examinations due unsatisfactory images and the possibility of post-acquisition digital image magnification with no need of an additional exposure of the patient for a magnification supplement^(20,21). Additionally, the digital detector has higher contrast dynamic range, allowing the use of a rhodium/rhodium filter/target combination in a higher number of images without quality degradation (Figure 1). Some authors have reported that an extra reduction of the dose also is determined by a higher digital detector quantum efficiency when compared to the intensifying screen-film system⁽²²⁾.

The detector performance may be evaluated by means of the measure modulation transfer function (MTF) and DQE. The MTF quantifies the resolution rate or image conspicuity and the DQE is a measure of the signal/noise ratio, responsible by the contrast resolution and dose efficiency. The image performance in radiography is characterized by the analysis of the DQE and MTF curves. However, a detector performance cannot be adequately described by a simple spatial frequency. These parameters are utilized to determine the best system for capturing data on a spatial frequency band. With a high MTF in a determined spatial frequency, small objects may be missed due to system noises. Increasing the signal and reducing the noises in the system, we increase the detection of small structures. The DQE, measure of the system signal/noise ratio in a spatial frequency, is a good measure of the dose efficiency. Several factors affect the DQE, including the quantity of absorbed X-rays, the amplitude or intensity of the signal profile (MTF measure) and noise.

It has been demonstrated that the full-field digital detector physical features present favorable results as to spatial resolution and DQE, when compared to the screen-film system^(5,6), and this may be particularly observed in low-contrast images of lesions^(23,24). Also, a better result of contrast detail was demonstrated⁽⁷⁾.

The minimum pixel size on the detector required for digital mammography has been subject for debate. The size of the pixel element on currently available detectors ranges between 50 µm and 100 µm. It has been observed that the reduction of the pixel size from 100 µm to 50 µm does not result in improvement of the observer performance in lesions characterization⁽²⁵⁾.

The separation of images acquisition, display and storage is another advantage of the digital mammography since it allows the image is electronically stored, being retrieved, displayed, manipulated and stored when and where necessary⁽¹⁶⁾.

The first digital system to be approved by the FDA, in January/2000, was the Senographe 2000D GE Medical Systems full-field digital mammography system with cesium iodide detector, followed by the Fischer Imaging Senoscan, in September/2001, the Selenia Hologic/Lorad digital mammograph with amorphous selenium detector, in March/2002 and the Siemens Mammomat Novation mammograph in October/2005⁽³⁾. Other companies, like Fuji, Sectra and Instrumentarium, had already their systems approved for use in other countries although their FDA approval processes have not be concluded yet.

TYPES OF FULL-FIELD DIGITAL MAMMOGRAPHY DETECTOR SYSTEMS

Phosphor plate system

The Senographe 2000D GE Medical Systems is an example of mammography system utilizing phosphor plate. In this system, there is a photodiodes matrix with an amorphous silicon substrate coupled with a cesium iodide phosphor plate. Each light-sensitive diode element is connected with a control line and a data line through a thin film transistor (TFT) in a way that the charge produced in a diode in response to the phosphor light emission is read and digitized. The detector pixel element size is approximately 100 µm, and the digitalization is of approximately 14 bits/pixel⁽²⁶⁾.

Phosphor-CCD system

The Senoscan Fischer Imaging is an example of this system. Cesium iodide phosphor, fiberoptics and charge-coupled device (CCD). The detector pixel element size is approximately 54 µm, and the digitalization is of approximately 12 bits/pixel⁽²⁷⁾.

Selenium system

It is a system different from those formerly described for not utilizing phosphor. A selenium photoconductor (Hologic/Lorad Selenia Digital Mammography System) absorbs X-rays that are directly converted into electronic signals, without the intermediate phase of X-rays conversion into light. Under the influence of an external electrical field, electrons float towards an electrode pixel and are collected in pixel capacitors. The detector pixel element size is approximately 70 µm, and the digitalization is of approximately 14 bits/pixel.

There are still few data available in the literature about this detector, but a potential advantage of this system is the high MTF and DQE.

CR system

It is an indirect conversion system not approved by the FDA yet, but it has been utilized in some European countries and Japan. The mammograph of Fuji Medical Systems is an example of this system. In the CR technology, the detector is a flexible plastic sheet coated with a photostimulable X-ray absorbing phosphor material. Imaging plates are located in cassettes for exposure in standard screen-film trays. In response to absorption of X-rays, electronic charges are stored in the phosphor crystalline material, where they remain stable for some time. After exposure, the image is read by a laser beam scanner. The laser light discharges the stored charge, causing emission of blue light that is collected by a light guide and detected by a photomultiplier tube. The resulting signal is logarithmically amplified, digitized and processed for film or soft-copy display. The pixel size of the resulting image is 50µm, with a digitalization accuracy of approximately 10 bits/pixel, after logarithmic compression.

The potential advantages of this system are: the smaller pixel size, the fact of utilizing a conventional mammographic X-ray system and its relatively low cost. It has been demonstrated that, despite the limitation caused by lower spatial resolution, the rate of microcalcifications detection by CR systems was equivalent to that of the conventional screen-film system^(9,17). Although the pixel size in the CR system is smaller than in the cesium iodide digital system (100 µm), the DQE is much lower. As a result, the CR system presents a higher noise rate compared with the direct acquisition system with the same radiation dose. Additionally, although both systems are limited by a five line pairs/mm spatial resolution, the direct acquisition system has a much higher resolution. This higher resolution is a result of the cesium iodide phosphor that produces images with higher resolution than the storage phosphor utilized in the CR systems.

FULL-FIELD DIGITAL MAMMOGRAPHY IMAGES FEATURES

Full-field digital mammography images may be displayed on laser-print films or on a computer screen.

Laser-print images offer a spatial resolution comparable to that of the conventional mammographic film (> 4,800 x 6,400 matrix size) and with the reproduced size combined with the acquisition resolution of current scanners (< 41 µm). The image contrast is similar to that of mammographic film, with a 3.5–4.0 maximum optical density. Artifacts or processing variations present in conventional mammographic films with simple emulsion do not represent a significant problem for laser-print digital mammography films.

The use of laser-print film increases the cost of the digital mammography and the image is displayed for interpretation in a single fixed format. An additional disadvantage of the film image is the loss of the inherent dynamic range from 12–14 bits to 8 bits⁽²⁷⁾.

Presently, only CRT (cathode ray tube) monitors are technologically able to display digital mammography images (Figure 2). Other technologies such as liquid crystal display, field emission display and organic light-emitting diodes may be available in the future^(28,29). However, the best CRT technology is limited in comparison with the laser-print film⁽²⁹⁾. The spatial resolution is lower than one-fourth of the film resolution and the luminance is much lower^(29,30).

Notwithstanding, these factors may be attenuated. A total spatial resolution is feasible by means of magnification (zoom) and contrast manipulation techniques (Figure 3). Also, the luminance difference may be irrelevant. Previous results have demonstrated that the mammographic findings detection rate does not decrease when the monitor luminance replaces the mammography negatoscope^(1,2,27).

However, researches including an evaluation of the monitor features effect on the breast cancer detection rate and diagnosis still remain necessary to assure both physicians and patients that print films and display images are equivalent. And besides, the interpretation speed in screening studies, with cost reduction should not be affected in case this technology is clinically implemented^(27,29).

DIGITAL MAMMOGRAPHY CLINICAL TRIALS

The first studies with clinical trials published in the literature were developed with equipment employing digital technology utilized in mammography-guided interventional procedures, i.e., utilizing small FOV detectors. Recently, the full-field digital mammography has been introduced in clinical trials for the purposes of comparison with the conventional mammography screen-film system^(5,18). Large clinical trials hitherto published are:

Colorado-Massachusetts Senographe 2000D screening trial^(31,32) – It was the first prospective clinical trial conducted for digital mammography in an asymptomatic population. This study involved 4,945 > 40 year old women who underwent screening mammography. Images reading was performed independently by different radiologists, each of them analyzing an equal number of conventional and digital mammography studies. The management of patients was made on the basis of findings in both full-field digital mammography and screen-film mammography systems.

In this study, conducted over a 30-month period, the final result being later published also by Lewin et al.⁽³²⁾ already including 6,736 pairs of digital and conventional mammography films of 4,489 patients, 179 biopsies were performed and cancer was diagnosed in 50 cases (42 image-aided and 8 interval cancers) There was no statistically significant difference in sensitivity or in the area under the ROC curve between the two systems. However, digital mammography resulted in fewer recalls than did the conventional mammography^(31,32). The workstation prototype utilized may have influenced these results since it did not allow manipulation of images by means of processing algorithms existing in more modern workstations, which may partially explain the lower cancer detection rate in the digital equipment utilized in the present study.

Norwegian Senographe 2000D trial-Oslo I and Oslo II study^(33,34) – The first phase of this study has evaluated 3,683 asymptomatic and symptomatic women by means of conventional and digital mammography⁽³³⁾. The authors have concluded that, in the population evaluated, conventional and digital mammography are comparable in terms of cancer detection. They also have observed that there was partiality in their study, since the mammography readers were quite experienced in conventional mammography reading in comparison with their experience in digital images reading on the workstation monitor, and the reading conditions for digital mammography were not the ideal ones, with many interruptions and excessive room light, particularly in the beginning of the study.

On the other hand, in the second phase of this study⁽³⁴⁾ randomized screening mammography was performed in 25,263 patients (70% underwent conventional mammography and 30% digital mammography). In this phase, the image reading was performed in a dedicated, dark room, with no interferences. The interpreters had already gained experience in digital images reading in workstations. For analyses results purposes, the patients were divided into two groups, one including women in the age group between 45 and 49 years, and the other, between 50 and 69 years. The cancer detection rate with the digital system was higher in patients in the second age group (0.83% with the digital system and 0.54% with the conventional system) and this difference has come near the statistical significance, whereas in patients between 45 and 49 years, there was no statistically significant difference in the cancer detection rate between the two systems (0.27% with the digital system and 0.22% with the conventional system). Differently from the study developed by Lewin et al.⁽³²⁾, where the recall rate was lower in the digital mammography than in the conventional mammography, the present study results have shown a significant increase in the recall rate in the digital mammography in relation to the conventional mammography. This may be partially explained by the higher rate of detected cancers in the digital equipment and also because, in the United States, the mammographic screening recall rate is higher than in Norway due to medical-legal implications.

American College of Radiology Imaging Network (ACRIN) DMIST^(35,36) – Other large clinical trial conducted for screening with digital mammography in an asymptomatic population, is the Digital Mammographic Imaging Screening Trial (DMIST), that has completed its recruitment target November, 2003. The initial funding granted for ACRIN was of US$ 22 million for a five-year period (1999–2004 inclusive). Funding for ACRIN was renewed for an additional four-year period (up to 2008).

In October, 2001, ACRIN initiated the DMIST. The primary purpose of this study was to evaluate the diagnostic accuracy of the digital mammography versus conventional mammography in asymptomatic women enrolled for screening mammography. All the participating patients underwent both digital and conventional mammography. Two radiologists interpreted independently each image of the study. The abnormalities management proceeded if any of the studies resulted positive. The result on the existence or not of cancer was determined by means of biopsy or clinical follow-up including mammography one year after enrollment for the majority of women with negative results at the initial mammography.

The study utilized FFDM systems from five different manufacturers: Senographe 2000D GE Medical Systems, Senoscan Fischer Imaging, Selenia Hologic/Lorad, Digital Mammography System Trex Medical, and Computed Radiography System Fuji Medical Systems. A total of 33 centers in Canada and United States have been involved in this study, with 49,528 women evaluated.

The DMIST is an important study because has allowed researchers to more accurately estimate both the sensitivity and specificity of digital mammography versus conventional mammography, due to the high number of participating women, providing more information on systems performance in specific lesions evaluation (calcifications and nodules). The cost/benefit ratio of the digital mammography versus conventional mammography also has been evaluated in this study, besides the effect of the breast density on the diagnostic perception in digital versus conventional mammographs and the diagnostic accuracy of each of the mammoghraphic units utilized. The initial results from this study were presented in September, 2005⁽³⁶⁾ and demonstrated that in the whole study population the diagnostic accuracy was similar in both methods in the area under the ROC curve. However, the digital mammography accuracy was significantly higher than in the conventional mammographic film amongst women with less than 50 years of age (area under the ROC curve for digital mammography: 0.84 ± 0.03; area under ROC curve for conventional mammographic film: 0.69 ± 0.05; difference 0.15, 95% confidence interval: 0.05 to 0.25; p = 0.002), in women presenting heterogeneously or extremely dense breasts at mammography (area under the ROC curve for digital mammography: 0.78 ± 0.03; area under the ROC curve for conventional mammographic film: 0.68 ± 0.03; difference 0.11, 95% confidence interval: 0.04 to 0.18; p = 0.003) and in pre- and perimenopausal women (area under the ROC curve for digital mammography: 0.82 ± 0.03; area under the ROC curve for conventional mammographic film: 0.67 ± 0.05; difference 0.15, 95% confidence interval: 0.05 to 0.24; p = 0.002). There was no statistically significant difference in the area under the ROC curve between digital mammography and conventional film mammographic amongst women aged 50 or more, women presenting entirely fat breast or scattered fibroglandular densities and postmenopausal women.

DIGITAL MAMMOGRAPHY CLINICAL ADVANTAGES

Many digital mammography benefits perceived by those who first have utilized this method have not been reported in the previously mentioned literature. This includes both operational and actual advantages from the diagnostic ability. Although not being easily measurable, these benefits should positively affect patients and their physicians.

Digital mammography, as well as other digital imaging methods, allows storage and transmission of each study, eliminating films loss and, eventually, the necessity of film files. The images may be electronically transmitted simultaneously to several physicians, or provided to patients without image quality loss. This is an important operational change. Presently, however, images cannot be electronically transmitted yet between institutions due to incompatibility between systems and for network security reasons. Besides, digital mammography images constitute quite heavy files, demanding high transmission speed, so centers still provide patients with images hard-copies.

Even more important for mammography than the digital image operational benefits, is the elimination of artifacts like dust and structural noises caused by the film processing. The digital mammography also reduces the variability of contrast, density, radiation dose and exposure time associated with film emulsion and processing. The film processing, particularly, is the greatest cause of images variation and requires daily checking of quality control parameters and variations correction. Insufficient contrast is a problem inexistent in digital images, as well as dust artifacts, because the detector is sealed.

Amongst the improvement prospects for the patient, the greatest advantage of the digital mammography is the quickness. This is especially true in diagnostic studies where the radiologist analyses each image before deciding the next step for the diagnostic evaluation. Because the patient does not need to wait for films developing, the mammographic examination takes much less time to be completed. Procedures for preoperative wire localization are even more affected, considering that, for this procedure, the patients must remain under compression while the last image is acquired and processed and then evaluated by the radiologist. As there is a reduction to 10–20 seconds of the time between exposure and image display (while, for conventional film processing, this time is of about three minutes), there is an important decrease both in the time required for the examination and the patient discomfort. The decrease in the time for examination also brings advantages to physicians, technicians and especially to radiological clinics administrators, due to efficiencies in terms of costs with facilities and technicians wages as examinations may be scheduled with smaller intervals in digital units. In many cases, these savings do not cover the high cost of a digital equipment, but the reduction of costs with films and archives, partially justifies the adoption of the digital mammography.

Although there is no expectation in terms of microcalcifications detection in mammographic screening, the digital mammography is ideal for microcalcifications characterization either by means of digital magnification of screening images or acquisition of high resolution magnified images with fine focus device, due the images low noise, besides the capability of manipulating a wide range of exposures through the workstation postprocessing resources^(9,16,17). Besides low noise, also there is the advantage of the wide dynamic range and high image contrast resulting from the acquisition made with higher voltage and less exposure time, reducing the patient mobility. The only disadvantage of the digital mammography is the lower spatial resolution that is resolved by digital geometrical magnification, increasing the projected image of calcifications above the spatial resolution lowest limit of the digital system (Figure 4). An issue precociously analyzed in this technology is that just the magnification of the image on the monitor is not a substitute for the true geometric magnification with fine focus. Although it was expected that the digital mammography could eliminate the necessity of patients recall for additional incidences with true geometric magnification, this has not occurred. In some cases, recall was necessary.

Figure 4 - click here to enlarge

Because of the wide contrast dynamic range, the digital mammography is ideal for showing implants images. A single exposure is sufficient for a clear demonstration of details of the implant itself or for optimization of the adjacent breast tissue visualization, provided adjustments are made in window parameters (Figure 5).

For the same reason, the digital mammography also is ideal for images of the skin and its immediately adjacent tissues. These tissues appear typically blackened on a well-exposed film of conventional mammography, requiring a special yellow light to partially retrieve information from that region. The image digital processing permits routine evaluation of the skin and adjacent tissues without using extra resources, just changing image window parameters. Although the skin usually is not relevant, there are diseases where it may be thickened, including the carcinoma (Figure 6). Additionally, tangential incidences may be performed in an attempt to investigate the cutaneous nature of some indeterminate calcifications, proving their benign character.

ADVANCED APPLICATIONS OF FULL-FIELD DIGITAL MAMMOGRAPHY

Computer-aided images analysis

The sensitivity of the mammographic screening has shown itself variable, probably due study models, selected population, technique employed, interval time between screening examinations and definition of terms like missed cancer, false-negative mammography, and interval cancer.

Missed cancer is defined as that where the biopsy has proved a cancer in an asymptomatic patient, with previously negative mammographic screening, but, retrospectively deemed visible⁽³⁷⁾. The term "missed" should not be used as an indication of negligence in the interpretation because the lesion evaluation was retrospectively performed. These cases should be considered as false-negative in auditing⁽³⁷⁾. Interval cancer is defined as that where a patient presents clinical findings before the next mammography of the screening program. This interval is usually of one year, but may vary, depending on the practice in the program site, and may reach 18–24 months.

Missed cancers may occur as a result of insufficient performance in lesions perception and analysis of the perceived lesions. Studies suggest that the decrease in the number of missed cancers occurs with the use of training methods, experience, continued education, prospective double reading, missed cases retrospective analysis and computer-aided detection (CAD) systems.

CAD programs were designed to provide fast visual commands to allow the radiologist to interpret more attentively specific areas of the image. Several programs different from CAD, for nodules and calcifications detection, have been developed^(38,39). There are many aspects affecting the CAD programs performance, including subtle lesions, size and localization, besides the features of the studies utilized in CAD training and the validation method utilized⁽³⁹⁾.

It has been suggested that the use of CAD programs in mammographic screening may result in an effectiveness increase, without an increment in time and workload employed to perform mammographic examinations⁽⁴⁰⁾. Detection rate was 84% of all the cases of cancer evaluated by CAD, with a 99% detection rate in cases of calcifications and 75% in nodules, all of them correctly marked⁽⁴⁰⁾. Also an increase in breast cancer early detection was evidenced, without increasing the number of patients recall for taking additional incidences or changing biopsies positive predictive value⁽⁴¹⁾.

All the currently available CAD systems produce marks indicating the areas that does not represent a breast cancer. With this system, reported rate of missed cancers in the screening was 77%, producing four marks, on average, in each mammographic study including four incidences, from which only one third presented missed cancers⁽⁴²⁾. The majority of marks indicate areas that the radiologist will opt to disregard due its benign appearance, being considered as false-positive. The efficiency and relative cost of CAD programs may be compared to the double reading performed by a same or by other radiologist⁽⁴²⁾.

The CAD program efficiency in the direct analysis of full-field digital mammography images has shown itself superior to that attained in the analysis of secondarily digitized images, resulting in 81% detection of calcifications and 81% detection of nodules⁽⁴³⁾. Other studies, however, have demonstrated equivalent rates for both methods, resulting in 97% detection of microcalcifications, 84% of nodules, with a 89% sensitivity⁽⁴⁴⁾.

The reproducibility of the marks produce by CAD systems is improving as a result of a decrease in the false-positive rates. The reproducibility of the identification of true-positive nodules remains as an issue with possible practical, methodological and clinical implications⁽⁴⁵⁾.

Contrast-enhanced digital mammography

Cancer is enhanced after intravenous iodine-based contrast medium injection. Digital detectors, due their capability to distinguish low-contrast structures and, using intravenous iodine-based contrast, identify tumors at a much lower cost than magnetic resonance imaging or computed tomography. This offers an additional opportunity to determine a mammographic finding relevance, as well as a potential improvement in the rate of early detection of breast cancer in a selected population.

By means of mask images, the minimum size of detectable vessels and optimum technical parameters required for breast digital subtraction angiography were defined and one has concluded that the images will allow identification of vessels < 0.20 mm in internal diameter, using a typical concentration for venous infusion. These results indicate that microvascularization associated with the tumor angiogenesis may be visualized with this method. With 20% of the radiological exposure dose usually employed in conventional mammography imaging, 0.25 mm vessels were identified, indicating the potential use of this dose in dynamic contrast-enhanced images⁽⁴⁶⁾.

Employing mammographic techniques of digital subtraction angiography, 18 patients with suspect images who underwent biopsy were evaluated. The conclusion is that the information provided by this method is qualitatively similar to those from magnetic resonance imaging with gadolinium-DTPA, but the mammography with digital subtraction angiography is faster and less cost-expensive than magnetic resonance imaging⁽⁴⁷⁾.

Digital tomosynthesis

Digital tomosynthesis is a method conceived more than thirty years ago⁽⁴⁸⁾, but could not be easily applied until the development of a digital detector that could be directly read, without the need to move the breast in the system. The development of this new method combines multiple X-ray pictures of each breast from many angles. The X-ray source moves on an arc around the breast, while the detector remains motionless^(49,50) (Figure 7). The images can be electronically reconstructed, allowing the characterization of different breast sectional planes. The total radiation dose for breast tomosynthesis is comparable to the dose of a single mammography incidence.

Because of the accurate knowledge on the X-ray tube in relation to the breast, the multiple images are aligned and overlapped, so just structures of sectional planes of interest may be aligned. Due to the lower radiation dose, the non-aligned planes present low contrast and low signal-to-noise ratio and a blurred appearance, while the signal of the aligned portions is added and the structures in the plane of choice become more visible.

A series of eight to ten images is obtained with the X-ray source moving through a 20–30º arc, effectively allowing few-millimeter-thick slices. A 3D model of the breast may be obtained with the use of reconstruction algorithms. This possibly should improve the capability of detecting tumors that currently are not seen due to the interface with structures behaving like noise.

Telemammography

Teleradiology offers a significant improvement in diagnostic efficiency and accuracy in relation to the current traditional screen-film-based diagnostic practice. The increase in the number of women requiring breast cancer screening makes the teleradiology advantages especially attractive for the digital mammography. On the other hand, the size and resolution of the digital mammography images represent great challenge for supporting a teleradiology effective cost.

Main applications of the telemammography include the possibility of creating a specialists center which could receive images from several sites for evaluation, allowing telediagnosis, teleconsultation and teleadministration⁽⁵¹⁾.

CONCLUSION

The results of the main current studies indicate a similar diagnostic accuracy between the digital mammography and the conventional mammographic film for breast cancer screening in the general population. Some results indicate an improvement in the detection rate for certain population segments (patients in determined age ranges and patients presenting dense breasts at mammography). Even if the digital mammography accuracy is not substantially different from the conventional mammographic film, the advanced applications available with the digital mammography, with the utilization of resources like CAD, contrast-enhanced mammography and digital tomosynthesis, show great promise to improve the breast cancer detection and early diagnosis.

REFERENCES

Received September 26, 2005.

Accepted after revision November 18, 2005.

• 1. Young KC, Wallis MG, Ramsdale ML. Mammographic film density and detection of small breast cancers. Clin Radiol 1994;49:461465.
• 2. Robson KJ, Kotre CJ, Faulkner K. The use of a contrast-detail test object in the optimization of optical density in mammography. Br J Radiol 1995;68:277282.
• 3. U.S. Food and Drug Administration. Center for Devices and Radiological Health. MOSA Program. Disponível em: URL: http://www.fda.gov/cdrh/mammography/digital.html
• 4. Shtern F. Digital mammography and related technologies: a perspective from the National Cancer Institute. Radiology 1992;183:629630.
• 5. Vedantham S, Karellas A, Suryanarayanan S, et al Full breast digital mammography with an amorphous silicon-based flat panel detector: physical characteristics of a clinical prototype. Med Phys 2000;27:558567.
• 6. Vedantham S, Karellas A, Suryanarayanan S, D'Orsi CJ, Hendrick RE. Breast imaging using an amorphous silicon-based full-field digital mammographic system: stability of a clinical prototype. J Digit Imaging 2000;13:191199.
• 7. Suryanarayanan S, Karellas A, Vedantham S, Ved H, Baker SP, D'Orsi CJ. Flat-panel digital mammography system: contrast detail comparison between screen-film radiographs and hard-copy images. Radiology 2002;225:801807.
• 8. Jarlman O, Samuelsson L, Braw M. Digital luminescence mammography. Early clinical experience. Acta Radiol 1991;32:110113.
• 9. Funke M, Hermann KP, Breiter N. A magnification mammography technique using storage phosphor plates: experimental investigations concerning spatial resolution and detection of microcalcifications. Fortschr Röntgenstr 1997;167: 174179.
• 10. Obenauer S, Hermann KP, Grabbe E. Dose reduction in full-field digital mammography: an anthropomorphic breast phantom study. Br J Radiol 2003;76:478482.
• 11. Obenauer S, Luftner-Nagel S, von Heyden D, Munzel U, Baum F, Grabbe E. Screen-film vs full-field digital mammography: image quality, detectability and characterization of lesions. Eur Radiol 2002;12:16971702.
• 12. Undrill PE, O'Kane AD, Gilbert FJ. A comparison of digital and screen-film mammography using quality control phantoms. Clin Radiol 2000; 55:782790.
• 13. Funke M, Breiter N, Hermann KP, Oestmann JW, Grabbe E. Storage phosphor direct magnification mammography in comparison with conventional screen-film mammography a phantom study. Br J Radiol 1998;71:528534.
• 14. Kheddache S, Thilander-Klang A, Lanhede B, et al Storage phosphor and film-screen mammography: performance with different mammographic techniques. Eur Radiol 1999;9:591597.
• 15. Funke M, Breiter N, Hermann KP, Oestmann JW, Grabbe E. Magnification survey and spot view mammography with a new microfocus X-ray unit: detail resolution and radiation exposure. Eur Radiol 1998;8:386390.
• 16. Rong XJ, Shaw CC, Johnston DA, et al Microcalcification detectability for four mammographic detectors: flat-panel, CCD, CR, and screen-film. Med Phys 2002;29:20522061.
• 17. Fischer U, Baum F, Obenauer S, et al Comparative study in patients with microcalcifications: full-field digital mammography vs screen-film mammography. Eur Radiol 2002;12:26792683.
• 18. Yaffe MJ, Nishikawa RM, Maidment ADA, Fenster A. Development of a digital mammography system. Proc SPIE 1988;914:182188.
• 19. Sickles E. Current status of digital mammography. Excerpta Medical Int. Congr. Series. Amsterdam: Elsevier. Presented at 7th Int. Congress Senol, 1993.
• 20. Pisano ED, Yaffe M. Digital mammography. Breast Dis 1998;10:127135.
• 21. Feig SA, Yaffe MJ. Current status of digital mammography. Semin Ultrasound CT MR 1996;5: 424443.
• 22. Gennaro G, di Maggio C, Bellan E. Full-field digital mammography and patient dose reduction. Presented at 14th European Congress Radiology, 2002.
• 23. Evans DS, Mackenzie A, Workman A, Care K, Great Britain. Comparison of charge coupled devices and amorphous silicon based technology for full field digital mammography. Presented at 13th European Congress Radiology, 2001.
• 24. Karellas A, Vedantham S, Levis I, et al Evaluation of a full-field clinical prototype flat panel imager for digital mammography. Presented at the 84th Scientific Assembly of the Radiological Society of North America, 1998.
• 25. Levy L, Muller M, Priday K, Rick A. Influence of pixel size on diagnostic accuracy with digital mammography. Presented at the 13th European Congress Radiology, 2001.
• 26. Pisano ED, Kuzmiak C, Koomen M, Cance W. What every surgical oncologist should know about digital mammography. Semin Surg Oncol 2001;20:181186.
• 27. Pisano ED, Cole EB, Kistner EO, et al Interpretation of digital mammograms: comparison of speed and accuracy of soft-copy versus printed-film display. Radiology 2002;223:483488.
• 28. Blume H, Roehrig H, Ji TL. High-resolution high-brightness CRT displays systems: update on the state of the art. SID Digest 1994;94:219222.
• 29. Hemminger BM, Dillon AW, Johnston RE, et al Effect of display luminance on the feature detection rates of masses in mammograms. Med Phys 1999;26:22662272.
• 30. Roehrig H, Krupinski E. Image quality of CRT displays and the effect of brightness of diagnosis of mammograms. J. Digit Imaging 1998;11:187188.
• 31. Lewin JM, Hendrick RE, D'Orsi CJ, et al Comparison of full-field digital mammography with screen-film mammography for cancer detection: results of 4,945 paired examinations. Radiology 2001;218:873880.
• 32. Lewin JM, D'Orsi CJ, Hendrick RE, et al Clinical comparison of full-field digital mammography and screen-film mammography for detection of breast cancer. AJR Am J Roentgenol 2002;179: 671677.
• 33. Skaane P, Young K, Skjennald A. Population-based mammography screening: comparison of screen-film and full-field digital mammography with soft-copy readingOslo I study. Radiology 2003;229:877884.
• 34. Skaane P, Skjennald A. Screen-film mammography versus full-field digital mammography with soft-copy reading: randomized trial in a population-based screening programthe Oslo II study. Radiology 2004;232:197204.
• 35. Pisano ED, Gatsonis CA, Yaffe MJ, et al American College of Radiology Imaging Network Digital Mammographic Imaging Screening Trial: objectives and methodology. Radiology 2005;236: 404412.
• 36. Pisano ED, Gatsonis CA, Hendrick E, et al Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 2005;353:17731783.
• 37. Bird RE, Wallace TW, Yankaskas BC. Analysis of cancers missed at screening mammography. Radiology 1992;184:613617.
• 38. Chan HP, Lo SCB, Sahiner B, Lam KL, Helvie MA. Computer-aided detection of mammographic microcalcifications: pattern recognition with an artificial neural network. Med Phys 1995;22:15551567.
• 39. Kallergi M, Carney GM, Gaviria J. Evaluating the performance of detection algorithms in digital mammography. Med Phys 1999;26:267275.
• 40. Warren Burhenne LJ, Wood SA, D'Orsi CJ, et al Potential contribution of computer-aided detection to the sensitivity of screening mammography. Radiology 2000;215:554562.
• 41. Freer TW, Ulissey MJ. Screening mammography with computer-aided detection: prospective study of 12.860 patients in a community breast center. Radiology 2001;220:781786.
• 42. Birdwell RL, Ikeda DM, O'Shaughnessy KF, Sickles EA. Mammographic characteristics of 115 missed cancers later detected with screening mammography and the potential utility of computer-aided detection. Radiology 2001;219:192202.
• 43. Baum F, Fischer U, Obenauer S, Muller S, Benali K, Grabbe E. Computer-assisted diagnosis (CAD) in direct full field digital mammography (FFDM): preliminary results. Presented at the European Congress Radiology, 2002.
• 44. O'Shaughnessy KF, Castellino RA, Muller S, Benali K. Computer-aided detection (CAD) on 90 biopsy-proven breast cancer cases acquired on a full-field digital mammography system. Presented at the 87th Scientific Assembly of the Radiological Society of North America, 2001.
• 45. Zheng B, Hardesty LA, Poller WR, Sumkin JH, Golla S. Mammography with computer-aided detection: reproducibility assessment initial experience. Radiology 2003;228:5862.
• 46. Niklason LT, Kopans DB, Hamberg LM, Venkatakrishnan V, Castleberry DE, Opsahl-Ong B. Digital subtraction angiography of the breast: technical parameters and minimum detectable vessel size for a full-field digital mammography system. Presented at the 83rd Scientific Assembly of the Radiological Society of North America, 1997.
• 47. Jong RA, Yaffe MJ, Skarpathiotakis M, et al Contrast-enhanced digital mammography: initial clinical experience. Radiology 2003;228:842850.
• 48. Miller ER, McCurry EM, Hruska B. An infinite number of laminagrams from a finite number of radiographs. Radiology 1971;98:249255.
• 49. Niklason LT, Christian BT, Whitman GJ. Improved visualization of breast lesions with digital tomosynthesis imaging. Presented at the 82nd Scientific Assembly of the Radiological Society of North America, 1996.
• 50. Niklason LT, Kopans DB, Hamberg LM. Digital breast imaging: tomosynthesis and digital subtraction mammography. Breast Dis 1998;10:151164.
• 51. Lou SL, Sickles EA, Huang HK, et al Full-field direct digital telemammography: technical components, study protocols and preliminary results. IEEE Trans Inf Technol Biomed 1997;1:270278.

*

Study developed at Department of Imaging Diagnosis, Universidade Federal de São Paulo-Escola Paulista de Medicina, São Paulo, SP, Brazil.

Publication Dates

• Publication in this collection
  26 Sept 2006
• Date of issue
  Aug 2006

History

• Received
  26 Sept 2005
• Accepted
  18 Nov 2005