Search

Purpose To compare screening mammography recall rate, cancer detection rate (CDR), and positive predictive values (PPVs) for digital mammography before and after radiologist experience with digital breast tomosynthesis (DBT). Materials and Methods This retrospective study was approved by the institutional review board and compliant with HIPAA. The authors reviewed screening mammography audit data obtained from 2009 to 2014, during which 108 276 digital mammographic examinations were performed (50 062 before and 58 214 after experience with DBT). Recall rate, CDR, PPV of positive screening result (PPV 1 ), PPV of biopsy recommendation (PPV 2 ), and PPV of biopsies performed (PPV 3 ) of digital mammography for six radiologists were compared before (2009-2011) and after (2012-2014) experience with DBT. Radiologists worked in both a community setting, in which only digital mammography was available, and in two tertiary breast imaging centers, where they interpreted DBT images starting in 2012. Data were examined by using generalized linear mixed modeling wherein observations were nested for each radiologist over time. P < .05 was considered indicative of a statistically significant difference; 95% confidence intervals (CIs) were calculated. Results The average recall rate was 6.8% (range, 3.6%-9.7%) before experience with DBT and 7.9% (range, 5.5%-9.5%) after (P = .0316). Before experience with DBT, the recall rate increased only 0.01% for each year from 2009 to 2011 (P = .9727). After experience with DBT, the recall rate increased 0.65% for each year from 2012 to 2014 (P < .0127). CDR increased from 2.5 per 1000 examinations (95% CI: 2.2, 2.9) to 3.5 per 1000 examinations (95% CI: 3.0, 4.0; P = .0203). PPV 2 and PPV 3 increased significantly after experience with DBT, from 26.9% (95% CI: 19.9%, 35.3%) to 36.1% (95% CI: 31.7%, 40.7%; P = .0212) for PPV 2 and from 31.2% (95% CI: 24.0%, 39.3%) to 40.0% (95% CI: 35.5%, 44.6%; P = .0290) for PPV 3 . Conclusion Recall rate, CDR, PPV 2 , and PPV 3 of digital mammography increased after radiologist experience with DBT. © RSNA, 2017.

Purpose: To compare recall rate, types of abnormalities recalled, additional imaging required, biopsy positive predictive value (PPV), and cancer detection rate before and after implementation of screening digital breast tomosynthesis (DBT)., Materials and Methods: This retrospective analysis was approved by the institutional review board and complied with HIPAA. The requirement to obtain informed consent was waived. Results from all screening digital mammography (DM) examinations performed without tomosynthesis from March 1, 2011, through February 29, 2012, and DBT examinations performed from March 1, 2012, through February 28, 2013, were reviewed to identify all Breast Imaging Reporting and Data System (BI-RADS) category 0 examinations (needs additional imaging). Radiology and pathology reports were reviewed. The recall rate, biopsy PPV, and cancer detection rate were calculated. Statistical analysis was performed by using a two-proportions z test., Results: The recall rate was 9.3% (1175 of 12 577 examinations; 95% confidence interval [CI]: 8.8%, 9.9%) for DM and 6.4% (827 of 12 921 examinations; 95% CI: 6.0%, 6.8%) for DBT, an overall reduction of 31% (P < .00001). The recall rate was lower with DM than with DBT for masses (8.9% vs 26.8%, respectively), distortions (0.6% vs 5.3%), and calcifications (13.4% vs 20.3%) (P < .0001 for all). The recall rate was lower with DBT than with DM for asymmetries (13.3% vs 32.2%, respectively) and focal asymmetries (18.2% vs 32.2%) (P < .0001 for both). Diagnostic evaluation with ultrasonography (US) increased with DBT at the time of additional imaging (2.6% for DM vs 28.3% for DBT, P < .0001). There was no significant difference between DM and DBT with regard to biopsy PPV (30.2% vs 23.8%, P = .21) or cancer detection rate per 1000 patients (5.4 vs 4.6, P = .44)., Conclusion: With DBT, the recall rate decreased and the biopsy PPV and cancer detection rate did not decrease. The distribution of recalled abnormalities changed, and more patients were evaluated with US only., (© RSNA, 2014.)

Purpose: To determine whether the screening mammography recall rate for individual radiologists varies as a function of imaging site., Materials and Methods: Institutional review board-approved, HIPAA-compliant retrospective review of mammography audit data was performed between May 1, 2008, and September 1, 2011, for five radiologists with expertise in breast imaging interpreting mammograms at a community office practice and an academic referral hospital. Informed consent was waived. Both sites used full-field digital mammography and batch screening interpretation. Recall rates were calculated from audit data. Breast density, history of surgery or biopsy, and availability of prior mammograms were compared in a sample of 500 mammograms from each site by using χ(2) and two-sample t tests., Results: Radiologists interpreted a total of 74 297 screening mammograms between both sites during the study. The overall mean recall rate was 7.8% ± 0.2 (standard deviation). Overall recall rate at the community site was significantly lower than at the hospital site (6.9% ± 0.3 vs 8.6% ± 0.3, P < .001). Recall rates for each radiologist were also lower at the community site: 3.7% versus 6.5% (P < .001), 9.6% versus 12.0% (P < .001), 7.5% versus 9.1% (P = .01), 7.6% versus 11.5% (P < .001), and 5.6% versus 8.0% (P = .02). There was a significantly higher proportion of patients at the hospital site with a history of surgery (13.4% vs 5.6%, P < .001) and biopsy (7.0% vs 1.4%, P < .001) but no difference in the percentage with dense breasts or in the percentage with availability of prior mammograms. Mean patient age was lower at the hospital site: 56.1 versus 62.9 years (P < .001)., Conclusion: Recall rates were higher at the hospital site, probably primarily caused by patient population factors., (RSNA, 2013)

Purpose: To determine whether baseline screening breast magnetic resonance (MR) imaging studies have a higher rate of follow-up or biopsy recommendation than do studies with prior MR images available for comparison., Materials and Methods: This was an institutional review board-approved, HIPAA-compliant, retrospective study. Informed consent was waived. Reports from 650 consecutive screening breast MR imaging examinations performed in women between September 2007 and December 2008 were reviewed. All examinations were performed by using the same protocol, and images were interpreted by the same radiologists. Presence of comparison studies, Breast Imaging Reporting and Data System (BI-RADS) category, and biopsy results were recorded. Data were analyzed by using the χ(2) test, the two-sample test of proportions, and the Fisher exact test., Results: Mean patient age was 51 years (range, 25-81 years). Of the baseline studies, findings in 31 of 307 (10.1%) were interpreted as BI-RADS category 3 and findings in 18 of 307 (5.9%) were interpreted as BI-RADS category 4 or 5. Of the examinations with findings classified as BI-RADS category 4 or 5, the results in two of 18 (11.1%) were positive for malignancy at biopsy. Of the examinations with prior MR images for comparison, findings in nine of 343 (2.6%) were interpreted as BI-RADS category 3 and findings in 16 of 343 (4.7%) were interpreted as BI-RADS category 4 or 5. Of the examinations with findings classified as BI-RADS category 4 or 5, the results in three of 16 (18.8%) were positive for malignancy at biopsy. The difference in the number of BI-RADS category 3 interpretations between the two groups was significant (P < .001), but there was no significant difference in BI-RADS category 4 or 5 interpretations or positive predictive values., Conclusion: Baseline screening MR imaging was associated with a higher likelihood of recommendation for short-interval follow-up than was MR imaging with prior images for comparison., (© RSNA, 2011.)

Purpose: To retrospectively assess the sensitivity and specificity of ultrasonographic (US)-guided fine-needle aspiration (FNA) of axillary lymph nodes for preoperative staging of breast cancer across a range of primary tumor sizes, by using histologic findings as a reference standard., Materials and Methods: Institutional review board approval was obtained for this HIPAA-compliant study; informed consent was waived. US-guided FNA results in 74 patients with breast cancer (75 axillae) were compared with final pathologic results. Lymph nodes were classified as benign, indeterminate, or suspicious on the basis of US characteristics at retrospective review. US-guided FNA in the most suspicious node at US, or the largest node if all appeared benign, was performed. Final pathologic results (sentinel lymph node biopsy [SNB] or axillary lymph node dissection [ALND]) were compared with US and preoperative US-guided FNA results. Results were assessed according to tumor size. Sensitivity, specificity, and positive predictive value of US and US-guided FNA were calculated., Results: Primary tumor sizes were 0.3-12 cm (mean, 3 cm). Patient age range was 31-81 years (mean age, 51 years). Sensitivity of US-guided FNA for predicting positive results at ALND or SNB was 71%-75%. Specificity was 100%. Sensitivity of US-guided FNA increased with primary tumor size., Conclusion: US-guided FNA of axillary lymph nodes in patients with newly diagnosed breast cancer had a sensitivity that increased with increasing size of the primary tumor., (RSNA, 2007)

Purpose: To retrospectively determine sensitivity and specificity of computed tomography (CT) for the diagnosis of appendicitis in pregnant women with nontraumatic abdominal pain and retrospectively compare findings at CT and ultrasonography (US) in patients who underwent both examinations, with surgery or clinical follow-up as a reference standard., Materials and Methods: Institutional review board approval was obtained, and the study was HIPAA compliant. Informed consent was waived. Findings of 80 consecutive CT examinations performed in 78 pregnant women (mean age, 25.9 years; range, 17-43 years) for abdominal pain between September 2000 and October 2004 were compared with findings at prior US (n=52), surgery, and clinical follow-up. Sensitivity and specificity were calculated for the diagnosis of appendicitis. The average fetal radiation dose was 16 mGy (1.6 rad) (range, 4-45 mGy [4-4.5 rad])., Results: CT findings were normal in 51 examinations (64%) and abnormal in 29 (36%). Abnormal findings were appendicitis (n=13), urinary tract calculi (n=6), small-bowel obstruction (n=2), cholelithiasis (n=2), pyelonephritis (n=2), diaphragmatic hernia (n=1), cecal bascule (n=1), ileus (n=1), and metastatic lymphadenopathy (n=1). One surgically confirmed case of appendicitis was not detected at CT. For diagnosis of appendicitis, sensitivity of CT was 92% (12 of 13 examinations), specificity was 99% (66 of 67), and negative predictive value was 99% (66 of 67). Fifty-two CT studies were performed after US. US findings were normal in 46 patients (88%) and abnormal in six (12%). Abnormal findings were cholelithiasis (n=3), obstructive hydronephrosis (n=1), small-bowel dilatation (n=1), and appendicitis (n=1). Among 46 patients with normal US findings, CT findings were abnormal in 14, nine of whom required surgery. CT added important diagnostic information in 14 of 46 patients (30%)., Conclusion: CT findings established the diagnosis in 35% of examinations in pregnant women with abdominal pain (28 of 80), with a negative predictive value of 99% for appendicitis; when CT followed negative US findings, CT findings established the diagnosis in 30% of patients., (Copyright (c) RSNA, 2007.)

Purpose: To retrospectively evaluate interobserver variability between breast radiologists by using terminology of the fourth edition of the Breast Imaging Reporting and Data System (BI-RADS) to categorize lesions on mammograms and sonograms and to retrospectively determine the positive predictive value (PPV) of BI-RADS categories 4a, 4b, and 4c., Materials and Methods: Institutional review board approval was obtained; informed consent was not required. This study was HIPAA compliant. Ninety-four consecutive lesions in 91 women who underwent image-guided biopsy comprised 59 masses, 32 calcifications, and three masses with calcification. Five radiologists retrospectively reviewed these lesions. Each observer described each lesion with BI-RADS terminology and assigned a final BI-RADS category. Interobserver variability was assessed with the Cohen kappa statistic. A pathologic diagnosis was available for all 94 lesions; 30 (32%) were malignant and 64 (68%) were benign. Pathologic analysis of benign lesions was performed on tissue obtained with image-guided core-needle biopsy. In cases referred for excisional biopsy after needle biopsy because of atypia or discordance, final surgical pathologic analysis was used for correlation with imaging findings. PPV for category 4 or 5 lesions was determined for all readers combined., Results: For ultrasonographic (US) descriptors, substantial agreement was obtained for lesion orientation, shape, and boundary (kappa = 0.61, 0.66, and 0.69, respectively). Moderate agreement was obtained for lesion margin and posterior acoustic features (kappa = 0.40 for both). Fair agreement was obtained for lesion echo pattern (kappa = 0.29). For mammographic descriptors, moderate agreement was obtained for mass shape, mass margin, and calcification distribution (kappa = 0.48, 0.48, and 0.50, respectively). Fair agreement was obtained for calcification description (kappa = 0.32). Slight agreement was obtained for mass density (kappa = 0.18). Fair agreement was obtained for final assessment category (kappa = 0.28). PPVs of BI-RADS category 4 and 5 assignments were as follows: category 4a, six (6%) of 102; category 4b, 17 (15%) of 110; category 4c, 48 (53%) of 91; and category 5, 71 (91%) of 78., Conclusion: Interobserver agreement with the new BI-RADS terminology is good and validates the US lexicon. Subcategories 4a, 4b, and 4c are useful in predicting the likelihood of malignancy., ((c) RSNA, 2006.)

Purpose: To determine the cost savings of stereotactic core needle biopsy over open surgical biopsy in patient subgroups defined according to mammographic findings., Materials and Methods: From July 1992 through February 1995, stereotactic core needle biopsy was performed in 356 women with 405 nonpalpable breast lesions (254 were masses and 151 were calcification). Lesions were classified according to mammographic finding, size, and level of suspicion. Two hundred three lesions were classified as indeterminate, 166 as suspicious, and 36 as highly suspicious. Medicare reimbursements for 1995 were used to determine costs and cost savings., Results: Overall cost savings for stereotactic core biopsy over open surgical biopsy was $741 per case. Average cost savings per case was $807 for masses and $630 for calcifications. The greatest savings occurred in the cases of indeterminate masses, with an average of $856 saved per case. The least savings occurred in the cases of highly suspicious calcifications, with $446 saved per case., Conclusion: Cost savings of stereotactic core needle biopsy vary in subgroups of patients defined according to mammographic findings. Over-all savings will depend on the distribution of patients among these groups. In this series, cost savings were realized with stereotactic core biopsy over open surgical biopsy for all mammographic subgroups.

Purpose: To determine if lesion size or number of calcifications affects the ability to obtain microcalcifications or a specific histologic diagnosis at stereotaxic core needle biopsy (SCNB)., Materials and Methods: Mammographic findings and histopathologic reports of 138 lesions in 124 patients (aged 30-87 years; mean age, 56.2 years) who underwent SCNB of calcifications were reviewed. Calcifications in the specimen and attainment of a specific diagnosis were correlated with lesion size and number of calcifications., Results: Calcifications were obtained in 118 cases (86%). A specific diagnosis was reported in 72 cases (52%). Differences in retrieval of calcifications or ability to establish a specific diagnosis with decreasing lesion size or decreasing number of calcifications were not statistically significant. Attainment of a specific diagnosis was significantly related to retrieval of calcifications (P<.005)., Conclusion: SCNB was successful in obtaining calcifications in a high percentage of cases regardless of lesion size or number of calcifications. When calcifications were retrieved, a specific diagnosis was attained in most cases (72 of 118).