Evaluation of the predictive ability of ultrasound-based assessment of breast cancer using BI-RADS natural language reporting against commercial transcriptome-based tests.

PURPOSE: The objective of this study was to assess the classification capability of Breast Imaging Reporting and Data System (BI-RADS) ultrasound feature descriptors targeting established commercial transcriptomic gene signatures that guide management of breast cancer. MATERIALS AND METHODS: This retrospective, single-institution analysis of 219 patients involved two cohorts using one of two FDA approved transcriptome-based tests that were performed as part of the clinical care of breast cancer patients at Harbor-UCLA Medical Center between April 2008 and January 2013. BI-RADS descriptive terminology was collected from the corresponding ultrasound reports for each patient in conjunction with transcriptomic test results. Recursive partitioning and regression trees were used to test and validate classification of the two cohorts. RESULTS: The area under the curve (AUC) of the receiver operator curves (ROC) for the regression classifier between the two FDA approved tests and ultrasound features were 0.77 and 0.65, respectively; they employed the 'margins', 'retrotumoral', and 'internal echoes' feature descriptors. Notably, the 'retrotumoral' and mass 'margins' features were used in both classification trees. The identification of sonographic correlates of gene tests provides added value to the ultrasound exam without incurring additional procedures or testing. CONCLUSIONS: The predictive capability using structured language from diagnostic ultrasound reports (BI-RADS) was moderate for the two tests, and provides added value from ultrasound imaging without incurring any additional costs. Incorporation of additional measures, such as ultrasound contrast enhancement, with validation in larger, prospective studies may further substantiate these results and potentially demonstrate even greater predictive utility.

False-negative results of breast MR computer-aided evaluation in patients with breast cancer: correlation with clinicopathologic and radiologic factors.

PURPOSE: To investigate clinicopathologic and radiologic factors associated with false-negative results of magnetic resonance computer aided evaluation (MR-CAE) of breast MR imaging (MRI) in breast cancer patients. MATERIALS AND METHODS: A total of 135 breast cancer patients who underwent preoperative breast MR-CAE were included. False-negative and true-positive MR-CAE results were compared in terms of clinicopathologic and radiologic features. Univariate and multivariate analyses were used to evaluate independent factors associated with false-negative results. RESULT: Six of 135 cancer patients (4.4%) were false negative on MRI. CONCLUSION: Breast cancer with false-negative results on MR-CAE was classified as a lower BI-RADS category on breast ultrasound than true-positive results on MR-CAE.

Differentiating fat necrosis from recurrent malignancy in fat-grafted breasts: an imaging classification system to guide management.

BACKGROUND: In breast reconstruction with autologous fat grafting, concerns persist about the ability to differentiate palpable masses representing fat necrosis from recurrent cancer. The authors' objective was to develop standardized imaging classifications to distinguish benign from malignant lesions after fat grafting. METHODS: A database of 286 breast reconstruction patients undergoing fat grafting from 2006 to 2011 was retrospectively reviewed to identify patients with imaging of clinically palpable masses. All images were reviewed independently by a radiologist blinded to prior results. Lesions were classified, using the American College of Radiology Breast Imaging Reporting and Data System ultrasound lexicon, as follows: A, solid mass, hypoechoic; B, solid mass, isoechoic; C, solid mass, hyperechoic; D, solid mass, complex echogenicity; E, anechoic mass with posterior acoustic enhancement; F, cystic mass with internal echoes; and G, negative. Evolutions in lesions on follow-up ultrasound were recorded. Images were correlated with histopathologic results. RESULTS: On ultrasound, 66 lesions were visualized in 37 patients with palpable masses. Twenty-two lesions (33 percent) were Breast Imaging Reporting and Data System category 4 lesions; biopsies were performed on all of them. Histopathologic results revealed that 85.7 percent (six of seven) with classification D and 100 percent with classifications A, B, C, E, F, and G were fat necrosis. The one malignant lesion (classification D) exhibited vascularity and angular margins on ultrasound and was not in the location of fat injection. Negative predictive value of avascularity and circumscribed margins for malignancy was 100 percent. Follow-up ultrasound of 29 lesions at a median of 6.5 months revealed that no masses increased in size or developed vascularity. CONCLUSION: Ultrasound analysis, with a standardized classification system, is reliable at differentiating benign from malignant lesions after fat grafting in breast reconstruction.

Breast ultrasound image classification based on multiple-instance learning.

Breast ultrasound (BUS) image segmentation is a very difficult task due to poor image quality and speckle noise. In this paper, local features extracted from roughly segmented regions of interest (ROIs) are used to describe breast tumors. The roughly segmented ROI is viewed as a bag. And subregions of the ROI are considered as the instances of the bag. Multiple-instance learning (MIL) method is more suitable for classifying breast tumors using BUS images. However, due to the complexity of BUS images, traditional MIL method is not applicable. In this paper, a novel MIL method is proposed for solving such task. First, a self-organizing map is used to map the instance space to the concept space. Then, we use the distribution of the instances of each bag in the concept space to construct the bag feature vector. Finally, a support vector machine is employed for classifying the tumors. The experimental results show that the proposed method can achieve better performance: the accuracy is 0.9107 and the area under receiver operator characteristic curve is 0.96 (p < 0.005).

The accuracy of BI-RADS classification of breast ultrasound as a first-line imaging method.

PURPOSE: The aim was to evaluate the accuracy of BI-RADS categories 3 - 5 in breast ultrasound (US) as the first-line imaging method. MATERIALS AND METHODS: 5077 examinations of a consecutive, unselected and mixed collective of symptomatic and asymptomatic patients were performed. Of these examinations, 835 cases of BIRADS 3 - 5 could be analyzed. RESULTS: The PPV with respect to a malignant lesion for BI-RADS 3, 4, 5 was 0.03, 0.48, and 0.97, respectively. When BI-RADS 4 and 5 cases are considered to be suspicious, the ratio of benign to malignant findings corresponds to 1:1.8. Analyzing BIRADS 3 - 5 lesions, the sensitivity, specificity and accuracy are 0.92, 0.85, and 0.87, respectively. CONCLUSION: The data support the feasibility of US for discriminating malignant from benign findings corresponding to the ACR BI-RADS classification without excessively increasing the number of unnecessary biopsies.

Subcategorization of ultrasonographic BI-RADS category 4: positive predictive value and clinical factors affecting it.

The objective of this study was to evaluate the positive predictive value (PPV) in ultrasonographically (US)-detected breast lesions of BI-RADS category 4a, 4b and 4c and to find how various clinical factors influenced the PPV of category 4. A total of 2142 women with 2430 breast lesions diagnosed on US as BI-RADS category 4 and underwent biopsy were included. Among them, 452 (18.6%) were pathologically confirmed as malignancy. PPV of each US BI-RADS subcategory was 7.6% (149/1963) for category 4a, 37.8% (68/180) for category 4b and 81.9% (235/287) for category 4c. Several clinical factors were more significantly seen in malignancy of category 4a and 4b, whereas none of the factors showed significance in category 4c. Subcategorization of category 4 is a feasible method in predicting malignancy in which patients' age, lack of multiplicity and symptoms affected the PPV of category 4 lesions.

Estado actual del rol de la 3D en la patología mamaria / Current role of 3D ultrasound in breast pathology

Descripción de distintas líneas de investigación que se han venido realizando, en cuanto a técnicas de diagnóstico y tratamiento de patologías mamarias. Para graficar las posibilidades diagnósticas de la ultrasonografía 3D se presenta el caso de una paciente de 50 años a la que se realizaron estudios ecográficos y análisis en 3D...

Carcinoma ductal in situ: correlación entre la histología y la resonancia magnética de mama / Breast MRI-histologic correlation for ductal carcinoma in situ

Objetivo: Evaluar la concordancia de la resonancia magnética (RM) de mama con la histología en la valoración del tamaño y extensión del carcinoma ductal in situ puro (CDIS), y compararla con la de las técnicas convencionales (mamografía y ecografía). Material y métodos: Estudio retrospectivo de pacientes consecutivas con biopsia percutánea con resultado de CDIS. Se estimó el coeficiente de concordancia de correlación de Lin para cada uno de las 3 técnicas de imagen con la histología. Además, se revisó la concordancia con gráficos de Bland-Altman. Se evaluaron los cambios de conducta quirúrgica generados por la RM. Resultados: El grupo estudiado fue de 32 pacientes. En cuanto al tamaño tumoral, la concordancia fue superior en la RM (0,78; intervalo de confianza [IC] de 95%, 0,62–0,87) que en la mamografía (0,43; IC del 95%, 0,19–0,62) o en la ecografía (0,27; IC del 95%, 0,09–0,43). La RM sobrestimó el tamaño con un promedio de 3mm, mientras que la mamografía y la ecografía lo subestimaron en 9 y 18mm, respectivamente. La RM fue superior para la detección del multifocalidad o multicentricidad (7 casos) frente a la mamografía (3 casos) y a la ecografía (0 casos). Hubo 6 cambios de conducta quirúrgica correctos basados en los hallazgos de la RM. Conclusión: La RM de mama es superior a las técnicas convencionales en la valoración del tamaño de los CDIS. Además, detecta más casos de multifocalidad y multicentricidad, por lo que aconsejamos su utilización prequirúrgica en pacientes diagnosticadas de CDIS, especialmente en mamas densas (AU)

Objective: To evaluate the concordance between the breast MRI findings and the histologic findings for the size and extension of pure ductal carcinoma in situ (DCIS) and to compare this concordance with that of conventional techniques (mammography and ultrasonography). Material and methods: This is a retrospective study of consecutive patients diagnosed with DCIS after percutaneous biopsy. We estimated Lin's coefficient of concordance for the histologic findings with each of the three techniques. We also assessed concordance using Bland-Altman graphs. Finally, we determined the impact of the MRI findings on the surgical management of patients with DCIS. Results: A total of 32 patients were included in the study. Concordance between imaging and histology on tumor size was higher for MRI (0.78; 95%CI, 0.62–0.87) than for mammography (0.43; 95%CI, 0.19–0.62) or for ultrasonography (0.27; 95%CI, 0.09–0.43). MRI overestimated the size of DCIS by a mean of 3mm, whereas mammography and ultrasonography underestimated it by 9mm and 18mm, respectively. MRI detected multifocality and multicentricity (7 cases) better than mammography (3) or ultrasonography (0). The MRI findings correctly changed the surgical management in six patients. Conclusion: Breast MRI is better than conventional techniques for the evaluation of the size of DCIS. Breast MRI also detects more cases of multifocality and multicentricity. We recommend that all patients diagnosed with DCIS (especially those with dense breasts) undergo breast MRI prior to surgery (AU)

Reclassificacao ultra-sonográfica das imagens mamograficas categoria 0, segundo o sistema BI-RADStm / Ultrasonographic reclassification of category 0 mammographic images, according to the BI-RADStm system

A neoplasia maligna da mama é atualmente a principal causa de morte por câncer em mulheres. Apesar do maior conhecimento sobre o comportamento biológico da doença, do aprimoramento nas técnicas de detecção precoce e das novas combinações terapêuticas, esse ainda permanece como um dos maiores problemas de saúde pública no Brasil e em todo o mundo (COSTA E SILVA, 1987; GOMES,1992; 1ª REUNIÃO DE CONSENSO DE MASTOLOGIA, 1994; BRASIL-CONPREV, 2002).

Breast imaging reporting and data system (BI-RADS).

The Breast Imaging Reporting and Data System (BI-RADS) lexicon was developed by the American College of Radiology to standardize mammographic reporting. The BI-RADS lexicon defines terms to describe abnormalities on mammograms, and it defines final assessment categories that are predictive of the likelihood of malignancy. Although the lexicon is clinically useful and facilitates communication and research, there is still substantial interobserver variability in its application. Lexicons for breast sonography and breast MRI are in progress.

Papilomatosis juvenil mamaria / Juvenile mammary papillomatosis

La papilomatosis juvenil mamaria es una enfermedad proliferativa benigna de pacientes jóvenes, habitualmente menores de 30 años. La presentación clínica más frecuente es la existencia de un nódulo mamario elástico y móvil. Anatomo-patológicamente se caracteriza por presentar hiperplasia epitelial ductal, en ocasiones con atipia marcada, y numerosos quistes de diferente tamaño entre otros hallazgos. Se ha asociado con un aumento en la incidencia de cáncer de mama, tanto en la propia paciente como en la familia. Revisamos la bibliografía sobre el tema. Presentamos los hallazgos mamográficos y ecográficos de una mujer de 22 años de edad diagnosticada de papilomatosis mamaria juvenil (AU)

Toward a standardized breast ultrasound lexicon, BI-RADS: ultrasound.

The approach outlined above for describing and reporting sonographic features of breast masses represents only the initial step in the development of a comprehensive system to enhance the accurate identification, reporting, and analysis of sonographic abnormalities of the breast. Future revisions, with validation of interobserver consistency in application of these descriptors across multiple centers, with feedback from potential users in the breast imaging community, will undoubtedly expand the utility of this effort.

Imaging of breast masses.

The American College of Radiology Breast Imaging and Reporting Data System (ACR BI-RADS) defines a mass as a space-occupying lesion seen in at least two projections. This article focuses on the management of breast masses with mammography and ultrasonography. Recommended work-up pathways are modified for nonpalpable masses, palpable masses in women over the age of 30, and palpable masses in women under the age of 30. The suggested protocols are based on scientific peer-reviewed literature whenever possible, but when evidence based studies are not available, the author relies on expert opinion.

Differentiation of benign and malignant breast tumors by logistic regression and a classification tree using Doppler flow signals.

In breast examinations with Doppler, an increased flow is found in malignant tumors. With the relatively new color Doppler, we measured different flow values in 133 cancer patients and in 325 women with benign disease. These measurements were used to develop diagnostic rules. For the highly correlated flow values, we used a stepwise procedure to select a final logistic regression model and a tree-based approach, which is a different way of modeling. With both approaches we developed simple diagnostic rules of which the sensitivity and the specificity exceeded 90%. There are no differences between the two approaches concerning discriminative ability. As complex statistical modeling leads to an overoptimism in the assessment of the error rates, we applied sensitivity analysis, investigated the stability of the selected logistic regression model, and estimated the magnitude of the overoptimism of the diagnostic rules with resampling methods. The results indicate that the estimates of sensitivity and specificity are probably close to realistic values for a clinical setting.

Parenchymal patterns of the breast defined by real time ultrasound.

The association of mammographic parenchymal patterns of the breast with breast cancer risk has been studied extensively but there is little information about the distribution of different patterns in populations at different risks for breast cancer. Such information could be obtained if a risk-free method of breast examination were available that could be applied to the general population. We have evaluated real time ultrasound for this application by comparing the parenchymal pattern as assessed by mammography with the extent of echogenicity in the breast on ultrasound examination in 102 subjects. Subjects were examined by both methods, the mammographic and ultrasound images independently classified, and the proportion of the breast occupied by radiological density or ductal prominence compared with the extent of echogenic areas on ultrasound. These two methods of classifying mammographic parenchymal patterns were found to be strongly correlated. Real time ultrasound may therefore be useful in the epidemiological study of mammographic pattern and breast cancer risk.