Breast Imaging Reporting And Data System Descriptors Research Articles

Abstract PS3-16: Identifying the most relevant descriptors when evaluating ultrasound images in breast cancer diagnostics: A secondary analysis of an international multicenter trial

Abstract Background and objectives: The Breast Imaging Reporting and Data System (BI-RADS) provides a standardized way to describe ultrasound images in breast cancer diagnostics. However, there is little information which descriptors are most strongly associated with malignancy and to which extend the single descriptors (tissue composition, shape, orientation, margin of lesion, echo pattern, posterior features, and calcifications) should be considered for the final evaluation of risk of malignancy. Thus, we aimed to identify which BI-RADS descriptors are most strongly associated with malignancy when evaluating ultrasound images in breast cancer diagnostics. Methods: This multicenter, prospective trial took place at 11 trial sites in Austria, France, Germany, Japan, Netherlands, Portugal, and the US from February 2016 to March 2019. The trial enrolled 1288 women presenting with a lesion ≥0.5 and ≤5 cm in 2D B-mode ultrasound. The examiner conducted a routine 2D B-mode ultrasound examination and had additional standard information about the patients’ disease history and family history. The examiner described the ultrasound images according to BI-RADS. All patients underwent histopathological confirmation which was the gold standard against which the clinical examiner was compared. We performed univariate and multivariate analyses using descriptive statistics, Chi-Square test, and logistic regression to identify which image descriptors are associated with malignancy. Results: Histopathologic evaluation showed malignancy in 368 of 1288 lesions (28.6%). The descriptors most strongly associated with malignancy were spiculated margins (rate of malignancy 84.9%; 79 of 93), calcification (69.9%; 51 of 73), un-parallel orientation (65.9%; 187 of 284), angular margins (64.6%; 64 of 99), posterior shadowing (62.4%; 88 of 142), irregular shape (55.2%; 208 of 377), and indistinct margins (52.0%; 185 of 356). Different tissue compositions and echo patterns were least useful to distinguish between malign and benign lesions. Upon multivariate analysis, calcifications (OR 5.52; 95% CI 1.94-15.87) and posterior shadowing (OR 16.13; 95% CI 2.75-90.91) remained significantly (p<0.05) associated with malignancy. Conclusion: We identified which BI-RADS descriptors are most strongly associated with malignancy when evaluating ultrasound images in breast cancer diagnostics. Future research may look into providing not only a standardized image description but also a standardized final evaluation for the rate of malignancy with respect to the different predictive usefulness of the single descriptors. This may further standardize and objectify the risk evaluation in breast cancer diagnostics. Trial registration: NCT02638935 Table 1: Association of BI-RADS descriptors with final histopathologic resultsbenign pathologymalignant pathologyp-valuetissue compositionp<0.0001homogeneous background texture; fat —no. (%)185 (60.3)122 (39.74)homogeneous background texture; fibroglandular —no. (%)378 (77.3)111 (22.7)heterogeneous background texture —no. (%)356 (72.7)134 (27.4)shape of lesionp<0.0001oval —no. (%)659 (86.1)106 (13.8)round —no. (%)89 (62.7)53 (37.3)irregular —no. (%)169 (44.8)208 (55.2)orientation of lesionp<0.0001parallel —no. (%)806 (82.6)170 (17.42)not parallel —no. (%)97 (34.15)187 (65.9)margin of lesionp<0.0001circumcised —no. (%)644 (89.0)80 (11.0)indistinct margin —no. (%)171 (48.0)185 (52.0)angular margin —no. (%)35 (35.4)64 (64.6)microlobulated margin —no. (%)117 (60.0)78 (40.0)spiculated margin —no. (%)14 (15.1)79 (84.9)echo patternp=0.02anechoic —no. (%)7 (100)0 (0.0)hyperechoic —no. (%)30 (79.0)8 (21.0)complex cystic and solid —no. (%)52 (82.5)11 (17.5)hypoechoic —no. (%)645 (71.0)264 (29.0)isoechoic —no. (%)40 (78.4)11 (21.6)heterogeneous —no. (%)136 (64.8)74 (35.2)posterior featuresp<0.0001none —no. (%)590 (73.2)216 (26.8)enhancement —no. (%)249 (83.3)50 (16.7)shadowing —no. (%)53 (37.6)88 (62.4)combined pattern —no. (%)20 (58.8)14 (41.2)calcificationp<0.0001no calcification —no. (%)894 (73.8)317 (26.2)calcification —no. (%)22 (30.1)51 (69.9) Citation Format: André Pfob, Richard G. Barr, Volker Duda, Christopher Buesch, Joerg Heil, Michael Golatta. Identifying the most relevant descriptors when evaluating ultrasound images in breast cancer diagnostics: A secondary analysis of an international multicenter trial [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS3-16.

Abstract P1-15-15: Accuracy of MRI after neoadjuvant therapy for invasive lobular carcinoma of the breast

Abstract Background: Invasive lobular carcinoma of the breast (ILC) has higher rates of false negative imaging than invasive ductal carcinoma, and lower rates of neoadjuvant therapy (NAT) use. We evaluated the accuracy of Breast Imaging Reporting And Data System (BIRADS) findings on magnetic resonance imaging (MRI) after either neoadjuvant chemotherapy or endocrine therapy, and determined whether imaging change correlates with disease free survival. Methods: We queried a database of 674 ILC cases treated at UCSF from 1981-2017 and identified all patients treated with NAT. We reviewed MRI reports and recorded BIRADS descriptors of findings, maximal tumor diameter for mass or non-mass enhancement (NME), and subjective radiologist comments on progression or improvement. We used the t-test, chi-squared test, Pearson's correlation, and Kaplan Meier survival estimates to evaluate the accuracy of MRI after NAT compared to true tumor size on pathology, and the relationship between imaging change and disease free interval in Stata 14.2. Results: Of 136 patients with ILC treated with NAT, we included 101 women who had a post-treatment breast MRI report available. Of these, 58.4% received neoadjuvant chemotherapy, and 41.6% neoadjuvant endocrine therapy. After NAT, MRI findings were mass only in 43%, both mass/NME in 33%, NME only in 18%, and neither in 5%. Maximal diameter of mass on post-treatment MRI underestimated true size by a mean of 3.3 cm (range -3.6 to 15.3 cm). NME size on post-treatment MRI underestimated true size by a mean of 1.87 cm (range -7.2 to 9.7 cm). Mass size on MRI underestimated true size by ≥1 cm in 61.5% of cases; this size discrepancy was associated with increased positive margins (46.4% versus 20%, p=0.011). NME size on MRI underestimated true size by at ≥1 cm in 65.6%. The correlation coefficient between mass size on MRI and true size was 0.34 (p=0.0041), which increased to 0.67 (p<0.0001) when excluding those with associated NME. The correlation coefficient between NME size on MRI and true size was 0.28 (p=0.1239). Subjective progression on post-treatment MRI was associated with increased recurrence rates (80% versus 18.3%, p=0.001). In those with subjective improvement on MRI, there was a trend towards longer disease free interval (89% versus 73% disease free at 4 years, p=0.13). Table 1.Patient and tumor characteristics. Neoadjuvant chemotherapy (n=59)Neoadjuvant endocrine therapy (n=42)P valueMean age (yrs, 95% CI)53.6 (50.9-56.3)61.3 (58.4-64.1)0.0002Subtype 0.105ER+ PR+ HER2-29 (53.7%)23 (62.16%) ER+ PR- HER2-14 (25.9%)13 (35.14%) ER- PR- HER2-1 (1.85%)0 HER2+10 (18.5%)1 (2.7%) Grade 0.076114 (25%)16 (38.1%) 237 (66.1%)26 (61.9%) 35 (8.93%)0 Surgical stage <0.001I17 (28.81%)28 (66.67%) II26 (44.07%)6 (14.29%) III16 (27.12%)8 (19.1%) Mean follow-up time (yrs, 95% CI)5.6 (4.53-6.75)5.1 (3.84-6.27)0.48 Conclusions: Maximal tumor diameter on MRI after NAT in ILC vastly underestimates true tumor size. While these findings suggest using caution when using an MRI for surgical planning in patients with ILC, particularly if there is associated NME, the trend towards improved disease free survival in those with a subjective improvement is intriguing and suggests that MRI changes could become an early predictor of outcomes. Citation Format: Fahrner-Scott KE, Wong JM, Piper M, Ewing C, Alvarado M, Esserman LJ, Hylton N, Mukhtar RA. Accuracy of MRI after neoadjuvant therapy for invasive lobular carcinoma of the breast [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P1-15-15.

Diagnostic value of breast imaging reporting and data system combined with wire-guided localization biopsy for breast microcalcifications in impalpable breast cancer

Objective To explore the diagnosis value of X-ray breast imaging reporting and data system (BI-RADS) combined with wire-guided localization biopsy for breast microcalcifications in impalpable breast cancer. Methods 192 palpation negative patients with 205 microcalcification lesions were detected by mammography. All lesions were classified according to BI-RADS descriptors for calcification and were categorized by the BI-RADS. The patients with BI-RADS category 4a and above underwent X-ray positioning guide wire-guided biopsy and pathological diagnosis. Results In 205 microcalcification lesions, 74 (36.1%) were malignant lesions, 131 (63.9%) were less than benign lesions. The positive predictive value of malignant breast lesions in clustered, segmental, regional linear branching calcifications were higher [83.3% (5/6), 100.0% (11/11), 100.0% (1/1)], followed by clustered, linear, segmental, regional pleomorphic calcifications [55.9% (38/68), 50.0% (1/2), 40.0% (8/20) and 33.3% (4/12), respectively]. The positive predictive values of malignant in linear branching calcifications and pleomorphic calcifications were significantly higher than those of coarse heterogeneous calcifications, amorphous or indistinct calcifications (χ2 values were 34.44, 51.87, 16.71, 29.86, all P < 0.05). The linear branching calcification had the highest possibility. The proportions of malignant lesions in four different types of glands were extremely dense 40.5% (30/74), heterogeneously dense39.2% (29/74), scattered areas of fibroglandular density 10.8% (8/74) and fat 9.5% (7/74), respectively. Conclusions BI-RADS categorization for breast microcalcification lesions can improve the detection rate of impalpable breast cancer. Linear branching calcification has higher predictive value for malignant lesions. Dense breast is the risk factors of breast cancer, which should be attached great importance. Key words: Mammography; Microcalcification; Biopsy; Palpation; Breast neoplasms

Logistic LASSO regression for the diagnosis of breast cancer using clinical demographic data and the BI-RADS lexicon for ultrasonography.

PurposeThe aim of this study was to compare the performance of image analysis for predicting breast cancer using two distinct regression models and to evaluate the usefulness of incorporating clinical and demographic data (CDD) into the image analysis in order to improve the diagnosis of breast cancer.MethodsThis study included 139 solid masses from 139 patients who underwent a ultrasonography-guided core biopsy and had available CDD between June 2009 and April 2010. Three breast radiologists retrospectively reviewed 139 breast masses and described each lesion using the Breast Imaging Reporting and Data System (BI-RADS) lexicon. We applied and compared two regression methods-stepwise logistic (SL) regression and logistic least absolute shrinkage and selection operator (LASSO) regression-in which the BI-RADS descriptors and CDD were used as covariates. We investigated the performances of these regression methods and the agreement of radiologists in terms of test misclassification error and the area under the curve (AUC) of the tests.ResultsLogistic LASSO regression was superior (P<0.05) to SL regression, regardless of whether CDD was included in the covariates, in terms of test misclassification errors (0.234 vs. 0.253, without CDD; 0.196 vs. 0.258, with CDD) and AUC (0.785 vs. 0.759, without CDD; 0.873 vs. 0.735, with CDD). However, it was inferior (P<0.05) to the agreement of three radiologists in terms of test misclassification errors (0.234 vs. 0.168, without CDD; 0.196 vs. 0.088, with CDD) and the AUC without CDD (0.785 vs. 0.844, P<0.001), but was comparable to the AUC with CDD (0.873 vs. 0.880, P=0.141).ConclusionLogistic LASSO regression based on BI-RADS descriptors and CDD showed better performance than SL in predicting the presence of breast cancer. The use of CDD as a supplement to the BI-RADS descriptors significantly improved the prediction of breast cancer using logistic LASSO regression.

External validation of a publicly available computer assisted diagnostic tool for mammographic mass lesions with two high prevalence research datasets.

Lesions detected at mammography are described with a highly standardized terminology: the breast imaging-reporting and data system (BI-RADS) lexicon. Up to now, no validated semantic computer assisted classification algorithm exists to interactively link combinations of morphological descriptors from the lexicon to a probabilistic risk estimate of malignancy. The authors therefore aim at the external validation of the mammographic mass diagnosis (MMassDx) algorithm. A classification algorithm like MMassDx must perform well in a variety of clinical circumstances and in datasets that were not used to generate the algorithm in order to ultimately become accepted in clinical routine. The MMassDx algorithm uses a naïve Bayes network and calculates post-test probabilities of malignancy based on two distinct sets of variables, (a) BI-RADS descriptors and age ("descriptor model") and (b) BI-RADS descriptors, age, and BI-RADS assessment categories ("inclusive model"). The authors evaluate both the MMassDx (descriptor) and MMassDx (inclusive) models using two large publicly available datasets of mammographic mass lesions: the digital database for screening mammography (DDSM) dataset, which contains two subsets from the same examinations-a medio-lateral oblique (MLO) view and cranio-caudal (CC) view dataset-and the mammographic mass (MM) dataset. The DDSM contains 1220 mass lesions and the MM dataset contains 961 mass lesions. The authors evaluate discriminative performance using area under the receiver-operating-characteristic curve (AUC) and compare this to the BI-RADS assessment categories alone (i.e., the clinical performance) using the DeLong method. The authors also evaluate whether assigned probabilistic risk estimates reflect the lesions' true risk of malignancy using calibration curves. The authors demonstrate that the MMassDx algorithms show good discriminatory performance. AUC for the MMassDx (descriptor) model in the DDSM data is 0.876/0.895 (MLO/CC view) and AUC for the MMassDx (inclusive) model in the DDSM data is 0.891/0.900 (MLO/CC view). AUC for the MMassDx (descriptor) model in the MM data is 0.862 and AUC for the MMassDx (inclusive) model in the MM data is 0.900. In all scenarios, MMassDx performs significantly better than clinical performance, P < 0.05 each. The authors furthermore demonstrate that the MMassDx algorithm systematically underestimates the risk of malignancy in the DDSM and MM datasets, especially when low probabilities of malignancy are assigned. The authors' results reveal that the MMassDx algorithms have good discriminatory performance but less accurate calibration when tested on two independent validation datasets. Improvement in calibration and testing in a prospective clinical population will be important steps in the pursuit of translation of these algorithms to the clinic.

A Comparison of Logistic Regression Analysis and an Artificial Neural Network Using the BI-RADS Lexicon for Ultrasonography in Conjunction with Introbserver Variability

To determine which Breast Imaging Reporting and Data System (BI-RADS) descriptors for ultrasound are predictors for breast cancer using logistic regression (LR) analysis in conjunction with interobserver variability between breast radiologists, and to compare the performance of artificial neural network (ANN) and LR models in differentiation of benign and malignant breast masses. Five breast radiologists retrospectively reviewed 140 breast masses and described each lesion using BI-RADS lexicon and categorized final assessments. Interobserver agreements between the observers were measured by kappa statistics. The radiologists' responses for BI-RADS were pooled. The data were divided randomly into train (n = 70) and test sets (n = 70). Using train set, optimal independent variables were determined by using LR analysis with forward stepwise selection. The LR and ANN models were constructed with the optimal independent variables and the biopsy results as dependent variable. Performances of the models and radiologists were evaluated on the test set using receiver-operating characteristic (ROC) analysis. Among BI-RADS descriptors, margin and boundary were determined as the predictors according to stepwise LR showing moderate interobserver agreement. Area under the ROC curves (AUC) for both of LR and ANN were 0.87 (95% CI, 0.77-0.94). AUCs for the five radiologists ranged 0.79-0.91. There was no significant difference in AUC values among the LR, ANN, and radiologists (p > 0.05). Margin and boundary were found as statistically significant predictors with good interobserver agreement. Use of the LR and ANN showed similar performance to that of the radiologists for differentiation of benign and malignant breast masses.

Scoring System Based on BI-RADS Lexicon to Predict Probability of Malignancy in Suspicious Microcalcifications

To develop a scoring system allowing quantification of the probability of malignancy of suspicious microcalcifications based on Breast Imaging Reporting and Data System (BI-RADS). A total of 163 microcalcification lesions surgically excised in 147 women (aged 28-65 years) were included. Two radiologists independently reviewed each lesion with BI-RADS (4th edition). The interobserver agreement and positive predictive value (PPV) for each descriptor were determined and multivariate analysis was used to develop a scoring system. The scores were compared between benign and malignant lesions and among BI-RADS categories by using the two-sample t-test or the analysis of variance. To assess the discriminative power of the scoring system, the area under the receiver-operating characteristic curve (AUC) and cutoff values for categorization were determined. For the test of the scoring system, the validation data set from a different facility was applied. Interobserver agreement was fair to moderate for distribution, morphology, and category (κ = 0.45, 0.40, and 0.37). PPVs were significantly different among BI-RADS descriptors and categories (P < 0.0001). Of the scoring system developed, the AUC was 0.75. The scores between benign and malignant lesions were significantly different. By means of cutoff values, PPV in category 4a, 4b, 4c, and 5 was 7.0%, 15.0%, 44.8%, and 83.3%, respectively. For the validation data set, the AUC was 0.79 and the PPV in category 4a, 4b, 4c, and 5 was 9.4%, 24.1%, 62.5%, and 80.0%, respectively. In suspicious microcalcifications, our scoring system based on BI-RADS (4th edition) could help to derive a specific final category with good stratification of the probability of malignancy.

Determination of Similarity Measures for Pairs of Mass Lesions on Mammograms by Use of BI-RADS Lesion Descriptors and Image Features

To determine similarity measures for selection of pathology-known similar images that would be useful for radiologists as a reference guide in the diagnosis of new breast lesions on mammograms. The images were obtained from the Digital Database for Screening Mammography developed by the University of South Florida. For determination and evaluation of similarity measures, the "gold standard" of similarities for 300 pairs of masses was determined by 10 breast radiologists. For determining similarity measures that would agree with radiologists' similarity determination, an artificial neural network (ANN) was trained with the radiologists' subjective similarity ratings and the image features. The image features were determined subjectively using the Breast Imaging Reporting and Data System (BI-RADS) lesion descriptors and objectively by computerized image analysis. The similarity measures determined by the ANN were compared to the gold standard and evaluated in terms of the correlation coefficient. The similarity measures determined using the BI-RADS descriptors only were not as useful as those determined by use of the image features only. When the BI-RADS margin ratings were combined with the image features, the correlation coefficient between the subjective ratings and the objective measures improved slightly (r = 0.76) compared to those based on the image features alone (r = 0.74). The inclusion of the BI-RADS margin descriptors may be useful for determination of similarity measures, especially when it is difficult to obtain the manual outlines of the masses and if the BI-RADS descriptors were provided consistently by radiologists.

Observer variability of Breast Imaging Reporting and Data System (BI-RADS) for breast ultrasound

PurposeTo evaluate inter- and intra-observer variabilities in breast sonographic feature analysis and management, using the fourth edition of the Breast Imaging Reporting and Data System (BI-RADS). Materials and methodsWe included 136 patients with 150 breast lesions who underwent breast ultrasound (US) and ultrasound-guided core needle biopsy. A pathological diagnosis was available for all 150 lesions: 77 (51%) malignant and 73 (49%) benign. Four radiologists retrospectively reviewed sonographic images of lesions twice within an 8-week interval. The observers described each lesion, using BI-RADS descriptors and final assessment. Inter- and intra-observer variabilities were assessed with Cohen's kappa statistic. Positive predictive value and negative predictive value (NPV) for final assessment were also calculated. ResultsFor inter-observer agreements for sonographic descriptors, substantial agreement for lesion calcification and final assessment (κ=0.61 for both), moderate agreement for lesion shape, orientation, boundary, and posterior acoustic features (κ=0.49, 0.56, 0.59, and 0.49, respectively), and fair agreement for lesion margin and echo pattern (κ=0.33 and 0.37, respectively) were obtained. For intra-observer agreement, substantial to perfect agreement was found for almost all lesion descriptors and final assessments. NPV for final assessment category 3 was 95%. Positive predictive value (PPV) for final assessment categorized as 4 or 5 were as follows: category 4a, 26%; category 4b, 89%; category 4c, 90%; and category 5, 97%. ConclusionBecause inter- and intra-observer agreement with the BI-RADS lexicon for US is good, the use of BI-RADS lexicon can provide accurate and consistent description and assessment for breast US.

Breast cancer CADx based on BI-RADS™ descriptors from two mammographic views

In this study we compared the performance of computer aided diagnosis (CADx) algorithms based on Breast Imaging Reporting And Data System (BI-RADS) descriptors from one or two views. To select cases for the study with different mediolateral (MLO) and craniocaudal (CC) view descriptors, we assessed the agreement in BI-RADS lesion descriptors, BI-RADS assessment, and subtlety ratings for 1626 cases from the Digital Database for Screening Mammogrpahy (DDSM) using kappa statistics. We used 115 mass caseswith different descriptors for the two views to design linear discriminant analysis (LDA) based CADx algorithms. The CADx algorithms used BI-RADS descriptors and patient age as features. Thealgorithms based on BI-RADS descriptors from both the views performed marginally betterthan algorithms based on BI-RADS descriptors from a single view. A system that averaged theresults of two classifiers trained separately on the MLO and CC views displayed the best performance (Az=0.920 +/- 0.027). Thus, some improvement in performance of BI-RADS based CADx algorithms may be achieved by combining information from two mammographic views.

Artificial neural network: improving the quality of breast biopsy recommendations.

To evaluate the performance and inter- and intraobserver variability of an artificial neural network (ANN) for predicting breast biopsy outcome. Five radiologists described 60 mammographically detected lesions with the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) nomenclature. A previously programmed ANN used the BI-RADS descriptors and patient histories to predict biopsy results. ANN predictive performance was compared with the clinical decision to perform biopsy. Inter- and intraobserver variability of radiologists' interpretations and ANN predictions were evaluated with Cohen kappa analysis. The ANN maintained 100% sensitivity (23 of 23 cancers) while improving the positive predictive value of biopsy results from 38% (23 of 60 lesions) to between 58% (23 of 40 lesions) and 66% (23 of 35 lesions; P < .001). Interobserver variability for interpretation of the lesions was significantly reduced by the ANN (P < .001); there was no statistically significant effect on nearly perfect intraobserver reproducibility. Use of an ANN with radiologists' descriptions of abnormal findings may improve interpretation of mammographic abnormalities.