Clinical Performance of Synthesized Two-dimensional Mammography Combined with Tomosynthesis in a Large Screening Population.

Purpose To compare the clinical performance of synthesized two-dimensional (s2D) mammography combined with digital breast tomosynthesis (DBT) with that of full-field digital mammography (FFDM) alone and FFDM combined with DBT in a large community-based screening population by analyzing recall rate, positive predictive value, and cancer detection rate. Materials and Methods This was a retrospective study approved by the institutional review board and was HIPAA compliant with waiver of informed consent. A total of 78 810 screening mammograms from October 11, 2011, to June 30, 2016, were retrospectively collected. Of these, 32 076 were FFDM, 30 561 were DBT-FFDM, and 16 173 were DBT-s2D mammograms. Diagnostic performance of FFDM, DBT-FFDM, and DBT-s2D mammography was compared. Statistical significance was determined by using the Pearson χ2 test and was expressed as odds ratios and related confidence intervals determined by means of logistic regression analysis with pairwise comparisons. Results Recall rates were significantly lower with DBT-s2D mammography (4.3%, 687 of 16 173) when compared with DBT-FFDM (5.8%, 1785 of 30 561; odds ratio, 0.72; 95% confidence interval: 0.65, 0.78; P < .0001) and when compared with FFDM alone (8.7%, 2799 of 32 076; odds ratio, 0.46; 95% confidence interval: 0.43, 0.51). The cancer detection rate was similar among FFDM alone (5.3 of 1000 screening examinations), DBT-FFDM (6.4 of 1000 screening examinations), and DBT-s2D mammography (6.1 of 1000 screening examinations) with no significant difference (FFDM vs DBT-FFDM, P = .08; FFDM vs DBT-s2D, P = .27). The percentage of invasive cancers detected was significantly higher with DBT-s2D mammography (76.5%) than with DBT-FFDM (61.3%, P = .01), and positive predictive values with DBT-s2D mammography (40.8%) were significantly higher than those with DBT-FFDM (28.5%, P < .0001). Conclusion Screening with DBT-s2D mammography in a large community-based practice improved recall rate and positive predictive values without loss of cancer detection rate when compared with DBT-FFDM and FFDM alone. © RSNA, 2017.

Utilizing digital breast tomosynthesis to improve accuracy of preoperative needle localization for surgical excisional biopsy.

We describe a case of an 88-year-old female who presented for needle localization to undergo excisional biopsy of a subtle asymmetry in the left breast, with successful localization achieved using digital breast tomosynthesis. Initial attempts at localization under 2D mammography were inaccurate. Subsequent digital breast tomosynthesis application for triangulation resulted in better visualization of the target, and successful localization. Specimen radiography confirmed the lesion was accurately targeted and pathology revealed ductal carcinoma in situ. Needle localization guided by mammography and inherent limitations of 2D mammography are discussed, along with a literature review of tomosynthesis guided needle localization.

Results of MR imaging screening for breast cancer in high-risk patients with lobular carcinoma in situ.

PURPOSE: To determine the outcome of screening breast magnetic resonance (MR) imaging examinations performed in patients with lobular carcinoma in situ (LCIS) at the authors' institution. MATERIALS AND METHODS: This study was approved by the institutional review board and was compliant with HIPAA. Retrospective review of screening breast MR imaging examinations at the institution from 1996 through September 2009 was performed in patients with prior biopsies demonstrating LCIS. Patients with prior breast cancer diagnosis were excluded. American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) recommendations were recorded. Pathologic results of all consequent biopsies were obtained from the electronic medical records. RESULTS: A total of 445 breast MR examinations in 198 patients with LCIS were identified. Of these, 308 were screening examinations in 134 patients. One patient was a BRCA mutation carrier and was excluded. Of the remaining 307 screening examinations, 254 (82.7%) had BI-RADS category 1 or 2 findings; 27 (8.8%) had BI-RADS category 3 findings; and 27 (8.8%) had B-IRADS category 4 or 5 findings. Of the 27 studies that led to a biopsy recommendation, 10 (37%) yielded benign pathologic findings, five (18.5%) yielded malignant pathologic findings, and seven (25.9%) yielded high-risk lesions. Of the 27 studies with BI-RADS 3 findings, two (7.4%) resulted in biopsy, findings of both were benign. Overall, malignancy was detected in five of 307 screening studies (1.6%) and in five of 133 screened patients (3.8%). The positive predictive value (PPV) of these screening studies for which biopsy was recommended was 18.5%. The PPV 3 (studies for which biopsy was recommended and actually performed, as described in the BI-RADS guidelines) was 23.8%. CONCLUSION: Screening breast MR imaging helped identify breast cancer in LCIS patients at a rate similar to that shown in high-risk populations for whom screening breast MR imaging is currently consistently recommended.

Auditing a breast MRI practice: performance measures for screening and diagnostic breast MRI.

PURPOSE: Breast MRI is increasingly used for both screening and diagnostic purposes. Although performance benchmarks for screening and diagnostic mammography have been published, performance benchmarks for breast MRI have yet to be established. The purpose of this study was to comprehensively evaluate breast MRI performance measures, stratified by screening and diagnostic indications, from a single academic institution. METHODS: Institutional review board approval was acquired for this HIPAA-compliant study. Informed consent was not required. Retrospective review of the institutional database identified all breast MRI examinations performed from April 1, 2007, to March 31, 2008. After application of exclusion criteria, the following performance measures for screening and diagnostic indications were calculated: cancer detection rate, positive predictive value (PPV), and abnormal interpretation rates. RESULTS: The study included 2,444 examinations, 1,313 for screening and 1,131 for diagnostic indications. The cancer detection rates were 14 per 1,000 screening breast MRI examinations and 47 per 1,000 diagnostic examinations (P < .00001). The abnormal interpretation rate was 12% (152 of 1,313) for screening and 17% (194 of 1,131) for diagnostic indications (P = .00008). The PPVs of MRI were lower for screening [PPV1 (abnormal findings) = 12%, PPV2 (biopsy recommended) = 24%, PPV3 (biopsy performed) = 27%] compared with diagnostic indications (PPV1 (abnormal findings) = 28%, PPV2 (biopsy recommended) = 36%, PPV3 (biopsy performed) = 38%]. CONCLUSIONS: Breast MRI performance measures differ significantly between screening and diagnostic MRI indications. Medical audits for breast MRI should calculate performance measures for screening and diagnostic breast MRI separately, as recommended for mammography.

A multichannel Markov random field framework for tumor segmentation with an application to classification of gene expression-based breast cancer recurrence risk.

We present a methodological framework for multichannel Markov random fields (MRFs). We show that conditional independence allows loopy belief propagation to solve a multichannel MRF as a single channel MRF. We use conditional mutual information to search for features that satisfy conditional independence assumptions. Using this framework we incorporate kinetic feature maps derived from breast dynamic contrast enhanced magnetic resonance imaging as observation channels in MRF for tumor segmentation. Our algorithm based on multichannel MRF achieves an receiver operating characteristic area under curve (AUC) of 0.97 for tumor segmentation when using a radiologist's manual delineation as ground truth. Single channel MRF based on the best feature chosen from the same pool of features as used by the multichannel MRF achieved a lower AUC of 0.89. We also present a comparison against the well established normalized cuts segmentation algorithm along with commonly used approaches for breast tumor segmentation including fuzzy C-means (FCM) and the more recent method of running FCM on enhancement variance features (FCM-VES). These previous methods give a lower AUC of 0.92, 0.88, and 0.60, respectively. Finally, we also investigate the role of superior segmentation in feature extraction and tumor characterization. Specifically, we examine the effect of improved segmentation on predicting the probability of breast cancer recurrence as determined by a validated tumor gene expression assay. We demonstrate that an support vector machine classifier trained on kinetic statistics extracted from tumors as segmented by our algorithm gives a significant improvement in distinguishing between women with high and low recurrence risk, giving an AUC of 0.88 as compared to 0.79, 0.76, 0.75, and 0.66 when using normalized cuts, single channel MRF, FCM, and FCM-VES, respectively, for segmentation.

Reader variability in breast density estimation from full-field digital mammograms: the effect of image postprocessing on relative and absolute measures.

RATIONALE AND OBJECTIVES: Mammographic breast density, a strong risk factor for breast cancer, may be measured as either a relative percentage of dense (ie, radiopaque) breast tissue or as an absolute area from either raw (ie, "for processing") or vendor postprocessed (ie, "for presentation") digital mammograms. Given the increasing interest in the incorporation of mammographic density in breast cancer risk assessment, the purpose of this study is to determine the inherent reader variability in breast density assessment from raw and vendor-processed digital mammograms, because inconsistent estimates could to lead to misclassification of an individual woman's risk for breast cancer. MATERIALS AND METHODS: Bilateral, mediolateral-oblique view, raw, and processed digital mammograms of 81 women were retrospectively collected for this study (N = 324 images). Mammographic percent density and absolute dense tissue area estimates for each image were obtained from two radiologists using a validated, interactive software tool. RESULTS: The variability of interreader agreement was not found to be affected by the image presentation style (ie, raw or processed, F-test: P > .5). Interreader estimates of relative and absolute breast density are strongly correlated (Pearson r > 0.84, P < .001) but systematically different (t-test, P < .001) between the two readers. CONCLUSION: Our results show that mammographic density may be assessed with equal reliability from either raw or vendor postprocessed images. Furthermore, our results suggest that the primary source of density variability comes from the subjectivity of the individual reader in assessing the absolute amount of dense tissue present in the breast, indicating the need to use standardized tools to mitigate this effect.

Breast MR imaging: normal anatomy.

Dynamic contrast-enhanced magnetic resonance (MR) imaging of the breast is a useful tool for the assessment of both structural and functional anatomy. A basic approach to the interpretation of normal anatomy on breast MR imaging is reviewed in this article.

Role of magnetic resonance imaging in evaluating the extent of disease.

Preoperative breast imaging evaluation can contribute useful clinical information to the management of the patient with known breast cancer. Breast magnetic resonance imaging (MRI) has been used as part of this imaging evaluation, and the ability of breast MRI to detect otherwise occult multifocal and multicentric disease has been demonstrated in multiple studies. The use of MRI for breast cancer staging remains under debate, however. This article reviews some of the current discussion regarding the use of breast MRI in this patient population. It is important to note that this discussion occurs in an evolving context of surgical and breast conservation therapies.