The purpose of this work was to evaluate computable Breast Imaging Reporting and Data System (BI-RADS) radiomic features to classify breast masses on ultrasound B-mode images. The database consisted of 206 consecutive lesions (144 benign and 62 malignant) proved by percutaneous biopsy in a prospective study approved by the local ethical committee. A radiologist manually delineated the contour of the lesions on greyscale images. We extracted the main ten radiomic features based on the BI-RADS lexicon and classified the lesions as benign or malignant using a bottom-up approach for five machine learning (ML) methods: multilayer perceptron (MLP), decision tree (DT), linear discriminant analysis (LDA), random forest (RF), and support vector machine (SVM). We performed a 10-fold cross validation for training and testing of all classifiers. Receiver operating characteristic (ROC) analysis was used for providing the area under the curve with 95% confidence intervals (CI). The classifier with the highest AUC at ROC analysis was SVM (AUC = 0.840, 95% CI 0.6667–0.9762), with 71.4% sensitivity (95% CI 0.6479–0.8616) and 76.9% specificity (95% CI 0.6148–0.8228). The best AUC for each method was 0.744 (95% CI 0.677–0.774) for DT, 0.818 (95% CI 0.6667–0.9444) for LDA, 0.811 (95% CI 0.710–0.892) for RF, and 0.806 (95% CI 0.677–0.839) for MLP. Lesion margin and orientation were the optimal features for all the machine learning methods. ML can aid the distinction between benign and malignant breast lesion on ultrasound images using quantified BI-RADS descriptors. SVM provided the highest ROC-AUC (0.840).
20 September 2018 in Journal of Immunology Research
Currently, attention has been given to complications related to breast implants, especially due to the presence of anaplastic large cell lymphoma (ALCL) related to silicone implants. Many manuscripts attempt to associate silicone presence with clinical complaints reported by patients, while others try to demonstrate the mechanisms of silicone bleeding by permeability loss of breast implant surfaces. There also are reports of foreign body type reactions from implant fibrous capsule to silicone corpuscles. However, there seems to be no study that correlates the clinical, radiological, and histological correlations of these lesions. The objective of this review is to correlate radiological findings of silicone-induced granuloma of breast implant capsule (SIGBIC) from breast MRI (BMRI) scans and complementary findings of ultrasound (US) and positron emission tomography (PET) scan, and its histology originated from surgical breast implant capsulectomy. To make this correlation possible, we divided SIGBIC into three radiological findings: (1) intracapsular SIGBIC, (2) SIGBIC with extracapsular extension, and (3) mixed SIGBIC associated with seroma. Our experience demonstrates histological-radiological correlation in SIGBIC diagnosis. Knowledge of these findings may demonstrate its real importance in terms of public health and patient management. We believe that SIGBIC is currently underdiagnosed by lack of training, guidance, and management in our clinical practice.
Complications resulting from the placement of silicone breast implants are becoming more frequent in our clinical practice. This is due to the increase in breast aesthetic surgeries at the beginning of the century, where breast augmentation using silicone implants was the main intervention performed. Generally, studies that discuss the complications of breast implants are restricted to reports of intra- or extra-capsular ruptures, contractures and haematomas. Currently, much importance has been given to anaplastic large cell lymphoma (ALCL) as a more severe complication related to silicone implants. Recently, granuloma formation induced by silicone particle bleeding from intact breast implants has been described when the free silicone comes into contact with the fibrous capsule of the implant. Few studies have demonstrated the characteristics and diagnostic keys for this entity. The objective of this study is to present cases of SIGBIC diagnosed in our service and to discuss the main findings that allow its diagnosis. Teaching Points • Breast implants induce fibrous capsule formation at the periphery of the implant. • Gel bleeding is inherent in all types of silicone breast implants. • Gel bleeding induces silicone-induced granuloma of breast implants. • Main diagnostic tips: heterogeneous mass, black-drop sign and late enhancement.
1 March 2017 in Breast Cancer: Targets and Therapy
Primary breast lymphoma is a rare disease and accounts for 0.5% of cases of breast cancer. Most primary breast lymphomas develop from B cells, and the involvement of T cells is rare. Anaplastic large cell lymphoma (ALCL) is a recently discovered T-cell lymphoma associated with breast implants. Only a few cases have been reported to date. It is believed that the incidence of ALCL is increasing because of the increasing number of breast implants. The clinical presentation is variable and can manifest as a palpable mass in the breast or armpit, breast pain, or capsular contracture. Because of the rarity of the disease and the lack of knowledge to date, clinical diagnosis is often delayed, with consequent delays in treatment. The cause and pathogenesis have not been fully elucidated, and there are no evidence-based guidelines for diagnosis, treatment, or follow-up of this disease. We present a review of cases of patients with silicone breast implants, including ALCL, a rare type of breast cancer that is still under study, and silicone-induced granuloma of breast implant capsule and its differential diagnosis, and discuss if a silicone-induced granuloma of breast implant capsule could be the precursor of the disease.
To evaluate the sensitivity (S) of BMRI to detect silicone gel bleeding in a prospective observational study, including consecutive patients referred for BMRI scan. From January 2017 to March 2018, we evaluated patients with breast implants referred for BMRI in a prospective observational study. For SIGBIC diagnosis, we adopted three new original imaging features: black drop signal; T2* hyper signal mass; and delayed contrast enhancement, considered as irrevocable signs to detect gel bleeding (GB). Histology confirmed the presence of a silicone corpuscle in breast implant capsular specimens. The accuracy of BMRI SIGBIC findings to predict GB was determined. We also compared SIGBIC diagnosis criteria to those features proposed by the BI-RADS léxicon, considered as equivocal findings. 208 patients had SIGBIC diagnosis at BMRI, and the histology confirmed GB in all cases. There were no false-positive results. Compared to the BI-RADS equivocal findings (S = 0.74), SIGBIC criteria had better sensitivity for GB diagnosis. SIGBIC diagnosis has high sensitivity to predict GB by the three irrevocable BMRI features described by the authors. We suppose GB is underdiagnosed in clinical practice by BI-RADS features. Study protocol: Plataforma Brasil CAAE: 77215317.0.0000.0072.