Search

Purpose The advent of three-dimensional breast imaging systems such as digital breast tomosynthesis (DBT) has great promise for improving the detection and diagnosis of breast cancer. With these new technologies comes an essential need for testing methods to assess the resultant image quality. Although randomized clinical trials are the gold standard for assessing image quality, phantom-based studies can provide a simpler and less burdensome approach. In this work, a complete framework is presented for task-based evaluation of microcalcification (MCs) detection performance for DBT imaging systems. Methods The framework consists of three parts. The first part is a realistic anthropomorphic physical breast phantom created through inkjet printing, with parchment paper and iodine-doped ink. The second is a method for inserting realistic MCs fabricated from calcium hydroxyapatite. The reproducibility and stability of the phantom materials were investigated through multiple samples of parchment and ink over 6 months. The final part is an analysis using a four-alternative forced choice (4AFC) reader study. To demonstrate the framework, a task-based 4AFC study was conducted using a clinical system to compare performance from DBT, synthetic mammography (SM), and full-field digital mammography (FFDM). Nine human observers read images containing MC clusters imaged with all three modalities and tried to correctly locate the MCs. The proportion correct (PC) was measured as the number of correctly detected clusters out of all trials. Results Overall, readers scored the highest with FFDM, (PC = 0.95 ± 0.03) then DBT (0.85 ± 0.04), and finally SM (0.44 ± 0.06). For the parchment and ink samples, the linear attenuation properties were very stable over 6 months. In addition, little difference was found between the various parchment and ink samples, indicating good reproducibility. Conclusions This framework presents a promising methodology for evaluating diagnostic task performance of clinical breast DBT systems.

Purpose Mammograms acquired with full-field digital mammography (FFDM) systems are provided in both “for-processing” and “for-presentation” image formats. For-presentation images are traditionally intended for visual assessment by the radiologists. In this study, we investigate the feasibility of using for-presentation images in computerized analysis and diagnosis of microcalcification (MC) lesions. Methods We make use of a set of 188 matched mammogram image pairs of MC lesions from 95 cases (biopsy proven), in which both for-presentation and for-processing images are provided for each lesion. We then analyze and characterize the MC lesions from for-presentation images and compare them with their counterparts in for processing images. Specifically, we consider three important aspects in computer-aided diagnosis (CAD) of MC lesions. First, we quantify each MC lesion with a set of 10 image features of clustered MCs and 12 textural features of the lesion area. Second, we assess the detectability of individual MCs in each lesion from the for-presentation images by a commonly used difference-of-Gaussians (DoG) detector. Finally, we study the diagnostic accuracy in discriminating between benign and malignant MC lesions from the for-presentation images by a pre-trained support vector machine (SVM) classifier. To accommodate the underlying background suppression and image enhancement in for-presentation images, a normalization procedure is applied. Results The quantitative image features of MC lesions from for-presentation images are highly consistent with that from for-processing images. The values of Pearson's correlation coefficient between features from the two formats range from 0.824 to 0.961 for the 12 MC image features, and from 0.871 to 0.963 for the 12 textural features. In detection of individual MCs, the FROC curve from for-presentation is similar to that from for-processing. In particular, at sensitivity level of 80%, the average number of false-positives (FPs) per image region is 9.55 for both for-presentation and for processing images. Finally, for classifying MC lesions as malignant or benign, the area under the ROC curve is 0.769 in for-presentation, compared to 0.761 in for-processing (p-value = 0.436). Conclusion The quantitative results demonstrate that MC lesions in for-presentation images are highly consistent with that in for-processing images in terms of image features, detectability of individual MCs, and classification accuracy between malignant and benign lesions. These results indicate that for-presentation images can be compatible with for-processing images for use in CAD algorithms for MC lesions. This article is protected by copyright. All rights reserved.

Purpose: Breast density is a significant breast cancer risk factor. Although various methods are used to estimate breast density, there is no standard measurement for this important factor. The authors are developing a breast density standardization method for use in full field digital mammography (FFDM). The approach calibrates for interpatient acquisition technique differences. The calibration produces a normalized breast density pixel value scale. The method relies on first generating a baseline (BL) calibration dataset, which required extensive phantom imaging. Standardizing prospective mammograms with calibration data generated in the past could introduce unanticipated error in the standardized output if the calibration dataset is no longer valid. Methods: Sample points from the BL calibration dataset were imaged approximately biweekly over an extended timeframe. These serial samples were used to evaluate the BL dataset reproducibility and quantify the serial calibration accuracy. The cumulative sum (Cusum) quality control method was used to evaluate the serial sampling. Results: There is considerable drift in the serial sample points from the BL calibration dataset that is x-ray beam dependent. Systematic deviation from the BL dataset caused significant calibration errors. This system drift was not captured with routine system quality controlmeasures. Cusum analysis indicated that the drift is a sign of system wear and eventual x-ray tube failure. Conclusions: The BL calibration dataset must be monitored and periodically updated, when necessary, to account for sustained system variations to maintain the calibration accuracy.

Automatic exposure control (AEC) is an important feature in mammography. It enables consistently optimal image exposure despite variations in tissue density and thickness, and user skill level. Full field digital mammography systems cannot employ conventional AEC methods because digital receptors fully absorb the x-ray beam. In this paper we describe an AEC procedure for slot scanning mammography. With slot scanning detectors, our approach uses a fast low-resolution and low-exposure prescan to acquire an image of the breast. Tube potential depends on breast thickness, and the prescan histogram provides the necessary information to calculate the required tube current. We validate our approach with simulated prescan images and phantom measurements. We achieve accurate exposure tracking with thickness and density, and expect this method of AEC to reduce retakes and improve workflow.

Contrast-detail experiments were performed to optimize technique factors for the detection of low-contrast lesions using a silicon diode array full-field digital mammography (FFDM) system under the conditions of a matched average glandular dose (AGD) for different techniques. Optimization was performed for compressed breast thickness from 2 to 8 cm. FFDM results were compared to screen-film mammography (SFM) at each breast thickness. Four contrast-detail (CD) images were acquired on a SFM unit with optimal techniques at 2, 4, 6, and 8 cm breast thicknesses. The AGD for each breast thickness was calculated based on half-value layer (HVL) and entrance exposure measurements on the SFM unit. A computer algorithm was developed and used to determine FFDM beam current (mAs) that matched AGD between FFDM and SFM at each thickness, while varying target, filter, and peak kilovoltage (kVp) across the full range available for the FFDM unit. CD images were then acquired on FFDM for kVp values from 23-35 for a molybdenum-molybdenum (Mo-Mo), 23-40 for a molybdenum-rhodium (Mo-Rh), and 25-49 for a rhodium-rhodium (Rh-Rh) target-filter under the constraint of matching the AGD from screen-film for each breast thickness (2, 4, 6, and 8 cm). CD images were scored independently for SFM and each FFDM technique by six readers. CD scores were analyzed to assess trends as a function of target-filter and kVp and were compared to SFM at each breast thickness. For 2 cm thick breasts, optimal FFDM CD scores occurred at the lowest possible kVp setting for each target-filter, with significant decreases in FFDM CD scores as kVp was increased under the constraint of matched AGD. For 2 cm breasts, optimal FFDM CD scores were not significantly different from SFM CD scores. For 4-8 cm breasts, optimum FFDM CD scores were superior to SFM CD scores. For 4 cm breasts, FFDM CD scores decreased as kVp increased for each target-filter combination. For 6 cm breasts, CD scores decreased slightly as kVp increased for Mo-Mo, but did not change significantly as a function of kVp for either Mo-Rh or Rh-Rh. For 8 cm breasts, Rh/Rh FFDM CD scores were superior to other target-filter combinations and increased significantly as kVp increased. These results indicate that low-contrast lesion detection was optimized for FFDM by using a softer x-ray beam for thin breasts and a harder x-ray beam for thick breasts, when AGD was kept constant for a given breast thickness. Under this constraint, optimum low-contrast lesion detection with FFDM was superior to that for SFM for all but the thinnest breasts.

Effects of x-ray scattering on full-field digital mammography are analyzed with the scattering model of Seibert and Boone [Med. Phys. 15, 567–575 (1988)]. A new method is introduced for the estimation of model parameters from measurements. It is shown that with breasts thinner than a certain threshold, removing the anti-scatter grid leads to an improved contrast-to-noise ratio with a smaller patient dose. A fast approximate algorithm is presented for determining the scattered field in a gridless digital mammogram.

Results of acceptance testing 18 full-field digital mammography systems for clinical use and of conducting annual physics surveys of 38 screen–film mammography systems were compared in terms of exposure times, mean glandular breast doses, and image quality. These evaluations were made using the same test tools on all systems, with emphasis on assessing automatic exposure control performance and image quality on both digital and screen–film systems using clinical techniques. Survey results indicated that digital mammography systems performed similarly to screen–film systems in terms of exposure times and mean glandular doses for thin to intermediate breasts, but that digital mammography systems selected shorter exposure times and lower mean glandular doses for thicker breasts. For all breast thicknesses, digital mammography systems yielded mean contrast–detail scores higher than those for screen–film systems. For all breast thicknesses, the 18 digital mammography systems demonstrated less variance in terms of exposure times, mean glandular doses, and contrast–detail scores than did the 38 screen–film systems tested. These results indicate that the clinical use of digital mammography may generally improve image quality for equal or lower breast doses, while providing tighter control on exposures and image quality than screen–film mammography.

This presentation will begin by comparing the image quality and dose characteristics of Stereotactic Breast Biopsy (SBB) with those of screen-film and full field digital mammography. Part 2 of the presentation will explore several practical aspects of the ACR SBB Accreditation Program. When SBB was introduced as a commercial product in the 1990's, x-ray breast imaging was limited to screen-film mammography. By the time SBB became common, all SBB systems utilized digital imaging technology. These small field of view SBB image receptors were the first common examples of digital imaging for the breast. The ACR is the only accrediting body that accredits SBB facilities. The ACR SBB QC Manual was published in 1999. Current methods of maintaining a QC program for SBB, current SBB equipment will be discussed. In addition, the process of an ACR Scheduled On-Site Survey (for facilities that have failed to meet ACR accreditation standards) will be presented. Learning Objectives: 1. Understand the image quality and dose characteristics of SBB, Screenfilm and full field digital mammography 2. Understand the current QC requirements for SBB facilities accredited by ACR 3. Understand the ACR process for Scheduled On-Site Surveys of facilities that have failed to meet the accreditation standards

Purpose: To investigate any change in the number of extra views acquired for screening mammography in transitioning from screen/film (SFM) to full field digital mammography (FFDM) and the resulting effect this had on total dose. Methods and Materials: This was a retrospective study of screening mammograms performed at our institution over a four year period. A total of almost 50,000 screening exams were performed during the last two years of SFM use and the first two years of FFDM implementation with a 24 × 29 cm detector. The picture archiving and communications system was used to view and record the number of images taken per screening procedure for all digital exams, whereas the radiology information system was the source for repeat analysis for the SF period. Results: Of the 23,516 patients who received screening mammograms on FFDM equipment, 36.5% had more than the standard four‐views (craniocaudal and mediolateral oblique views of each breast). There were 15.5% who had a total of 5 images, 12.1% had 6 images, 3.3% had 7 images, and 5.6% had 8 or more images. The total repeat rate averaged over all patients was 14.1%. In contrast, preliminary results show that the repeat rate for patients undergoing SFM screening for an equivalent 2 year period was close to 5.5%. As for dose, there was, on average, an 18% decrease in mean glandular dose per view in going from SFM to FFDM imaging. This dose saving is decreased considerably when the number of extra views is taken into consideration. Conclusions: A notable increase in the total repeat rate for screening mammograms was observed once the switch from SFM to FFDM occurred. These extra views had the undesired effect of offsetting some of the dose savings afforded by the digital equipment.

Full Field Digital Mammography (FFDM) systems are rapidly gaining in popularity and are replacing conventional film‐screen mammography (FSM) systems in breast centers throughout the country. Radiolgists are faced with unique challenges brought about by this transition of technology. In this lecture, the utility of full field digital mammography in clinical practice will be presented. Some of the challenges faced by radiologists who have transitioned from film screen mammography to digital mammography will also be discussed, and case examples illustrating the utility of FFDM will be presented. The clinical utility of computer‐aided detection (CAD) software packages will also be addressed. Learning objectives: 1. Understand the clinical impact that full‐field digital mammography has had on the daily practice of breast imaging. 2. Recognize the advantages and disadvantages of FFDM systems. 3. Understand some of the challenges that breast imagers face when transitioning from FSM to FFDM systems. 4. Understand how computer‐aided detection (CAD) systems are used in clinical practice.

Purpose: To design, implement, and evaluate a computer aided diagnosis(CADx) algorithm for the benign/malignant differentiation of breast calcification clusters that uses features defined from both breast views recorded with either screen/film (SF) or full field digital mammography (FFDM). Method and Materials: Two sets of data were tested in this work. An SF set of 101 pairs of images consisting of CC and MLO views of the same calcification cluster and an FFDM set of 50 pairs. Regions of interest, 512×512 pixels in size, were selected on each view and centered on the calcification cluster of interest. Eight different experiments were conducted with the SFM set and one experiment is in progress with the FFDM data. The SFM experiments involved different methods of feature estimation from the two views, different sets of features, as well as a single‐view based classification for comparison purposes. Results: On the single‐view run, the classifier achieved an AZ =0.889 with 88% sensitivity and 77% specificity at an operating point of 0.4; 12 features were selected as the most important. On the two‐view dataset, the classifier trained only with CC‐views achieved the highest AZ =0.968 and a sensitivity =98%; 9 features were the most important in this case. When the average value of features from the two views was used, classification achieved an AZ =0.958 with a sensitivity and specificity of 98% and 80% respectively. Combinations of features from the two views surpassed the average value classifier yielding an AZ =0.976 with 96% sensitivity and 82% specificity. Conclusion: The combination of two‐view information is instrumental in enhancing CADx performance but faces several challenges particularly in setting criteria for the optimum selection of features from the two views. Our method was transferred from film to digital without major modifications. FFDM results will be reported at the meeting.

Purpose: Our goal is to develop and evaluate breast tomosynthesis of amorphous selenium (a‐Se) based full field digital mammography (FFDM) system. The clinical value of tomosynthesis will be to provide additional clinical information in order to improve decision making accuracy to either confirm or exclude a suspected abnormality, verify correct target for biopsies and analyzetumor margin better to get knowledge of the possible extent breast cancer.Method and Materials: The prototype of tomosynthesis full field digital mammography system which is used in evaluation is based on amorphous selenium flat‐panel detector technology. The overall thickness of the selenium structure is 200μm, and the pixel size on this detector is 85μm. The total arc of the breast tomosynthesis system is 60° (−30° to +30°) and tomosynthesis sequence is performed at approximately 1–1.5 times the radiation dose of a conventional mammogram including 15 low‐dose exposures. Reconstruction methods will be also discussed. Results: Based on the experimental results of this study, we believe that FFDM tomosynthesis can be used to determine whether a mammography finding is caused by a real abnormal lesion or by superimposition of normal parenchymal structures to be able to diagnose and analyze the findings properly. Conclusion: We need to be able to find screening and diagnostic method, which has high sensitivity as well high specificity to avoid unnecessary surgical interventions and to be able to detect breast cancer in its early stage. The technology has already proved its benefits in detecting early breast cancers and the capability of diagnosing breast cancers. Although FFDM tomosynthesis evaluation needs to concentrate on proving that it is clinically successful in the sense of increased sensitivity and specificity, with lower cost of workflow and reduced risk. Conflict of Interest: Mari Varjonen and Pekka Strommer are employees of Planmed Oy.

Purpose: To develop a computer‐aided diagnosis(CAD) system for characterization of malignant and benign microcalcifications on mammograms acquired with a full field digital mammography (FFDM) system. Method and Materials: Our computerized classification system uses raw digital mammograms as input. The individual microcalcifications are first segmented by the system from the mammographic breast tissue background. Five morphological features describing the size, the density, and the shape of the individual microcalcifications are extracted. The mean and the variation of each of these features for the individual microcalcifications within a cluster are calculated. These cluster features in combination with the number of microcalcifications in a cluster form the morphological feature space. For texture feature extraction, a region of interest (ROI) containing the cluster of microcalcifications is identified on the mammogram. Background correction is applied to the ROI to reduce the intensity variation in the breast tissue areas. Texture features including the mean, entropy, contrast, and angular second moment are extracted from the gray level dependence difference statistics of the background‐corrected ROIs in four directions. A leave‐one‐case‐out resampling scheme is used to train and test the linear discriminant classifier. The most effective features from the combined morphological and texture feature space are identified using stepwise feature selection with simplex optimization in each training cycle. The performance of the classifier is evaluated by receiver operating characteristic (ROC) analysis. Results: In a preliminary study using digitized mammograms, the computer classifier obtained an area under the ROC curve, Az, of 0.82±0.03 for testing. The performance of the CAD system using a data set of 100 cases of biopsy‐proven microcalcifications on FFDMs will be presented. Conclusion: A trained CAD system can provide an estimate of the likelihood of malignancy of microcalcification on mammograms and thus may be used as a second opinion by radiologists for mammographic interpretation.