Search

Conventional deep learning (DL) algorithm requires full supervision of annotating the region of interest (ROI) that is laborious and often biased. We aimed to develop a weakly-supervised DL algorithm that diagnosis breast cancer at ultrasound without image annotation. Weakly-supervised DL algorithms were implemented with three networks (VGG16, ResNet34, and GoogLeNet) and trained using 1000 unannotated US images (500 benign and 500 malignant masses). Two sets of 200 images (100 benign and 100 malignant masses) were used for internal and external validation sets. For comparison with fully-supervised algorithms, ROI annotation was performed manually and automatically. Diagnostic performances were calculated as the area under the receiver operating characteristic curve (AUC). Using the class activation map, we determined how accurately the weakly-supervised DL algorithms localized the breast masses. For internal validation sets, the weakly-supervised DL algorithms achieved excellent diagnostic performances, with AUC values of 0.92-0.96, which were not statistically different (all Ps > 0.05) from those of fully-supervised DL algorithms with either manual or automated ROI annotation (AUC, 0.92-0.96). For external validation sets, the weakly-supervised DL algorithms achieved AUC values of 0.86-0.90, which were not statistically different (Ps > 0.05) or higher (P = 0.04, VGG16 with automated ROI annotation) from those of fully-supervised DL algorithms (AUC, 0.84-0.92). In internal and external validation sets, weakly-supervised algorithms could localize 100% of malignant masses, except for ResNet34 (98%). The weakly-supervised DL algorithms developed in the present study were feasible for US diagnosis of breast cancer with well-performing localization and differential diagnosis., (© 2021. The Author(s).)

Purpose: To investigate gray-scale and color Doppler sonographic (US) features of complex fibroadenoma (FA), according to the Breast Imaging-Reporting and Data System (BI-RADS) lexicon.Methods: From 2010 through 2013, 586 women with FAs were seen; 101 (17%) of those lesions were diagnosed as complex FAs on percutaneous needle biopsy or surgery. Among the patients with complex FAs, 67 who had US examination results available were included in this study. In addition, the results from 98 women who had simple FAs were included as controls. US features were retrospectively analyzed by two breast radiologists in consensus for shape, margin, echogenicity, posterior acoustic pattern, boundary, orientation, and associated findings. They also reassessed the BI-RADS category for FAs. Color Doppler US examination results were classified according to the amount of vascularity as absent, moderate, or marked.Results: Complex FA were larger than simple FAs were (14.5 cm versus 12.1 cm, p > 0.05). On univariate analysis, a round to irregular shape, an uncircumscribed margin, the presence of associated findings, and BI-RADS categorization as 4a and 4b were more frequently revealed in complex than in simple FAs (p < 0.05). Multivariate analysis revealed that the tumor margin was a predictive factor for complex FA (odds ratio: 6.08; 95% confidence interval: 1.14-32.49, p < 0.05). On color Doppler US, the complex FAs had higher degrees of vascularity than simple FA (p < 0.05).Conclusions: Complex FAs, in comparison with simple FAs, tend to have more aggressive features and to be in higher BI-RADS categories on US examination. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of our study was to evaluate the imaging features of benign adenomyoepithelioma of the breast with a focus on sonographic (US) appearances. Methods: Ten benign adenomyoepitheliomas in 9 patients were included in this study. Mammographic and US findings were analyzed retrospectively by 2 radiologists according to the Breast Imaging-Reporting and Data System (BI-RADS) lexicon. The BI-RADS final assessment category was also recorded. Results: Of the 8 lesions for which mammography was available, 4 lesions presented with abnormal findings. Two lesions presented with an indistinct, irregular, isodense mass and a circumscribed, oval, isodense mass. The other 2 lesions presented with areas of focal asymmetry and asymmetry, respectively. On US, all of 10 lesions presented as a mass. The masses measured 0.5-1.2 cm (mean, 0.8 cm), were irregular ( n = 8) or oval ( n = 2) with microlobulated ( n = 5), indistinct ( n = 3), or angular ( n = 2) margins. They were all hypoechoic, had non parallel orientation in 6 cases, and increased peripheral vascularity in 7 cases. The BI-RADS final assessment category was 4B in 6, and 4A in 4. Conclusions: Mammographic findings of benign adenomyoepithelioma were nonspecific. An irregular, non parallel orientation, microlobulated or indistinct margin, hypoechogenicity, and increased peripheral vascularity were the most frequent US features. © 2013 Wiley Periodicals, Inc. J Clin Ultrasound, 2013 [ABSTRACT FROM AUTHOR]

The purpose of this study was to determine whether disease recurrence and intrinsic characteristics of ductal carcinoma in situ (DCIS) are associated with the imaging method of detection in asymptomatic women. This multicenter, retrospective study included 844 women treated for asymptomatic DCIS who had pre-operative mammography and breast ultrasonography (US) studies available. Of the 844 women, 25 (3.0%) developed recurrences. Patients in the US group had significantly lower 5- and 10-y recurrence-free survival (RFS) rates compared with patients in the mammography group (p = 0.011). US-detected DCIS showed significantly lower 5-and 10-y RFS rates compared with mammography-detected DCIS in patients <50 y or with mammographically dense breasts (p = 0.002 and 0.002, respectively). US as the detection modality (hazard ratio [HR]: 4.451; 95% confidence interval [CI]: 1.530, 12.950; p = 0.006) and HER2 positivity (HR: 4.036; 95% CI: 1.438; 11.330; p = 0.008) were significantly associated with recurrence. We concluded that US as the detection modality and HER2 positivity were significantly associated with recurrence in patients treated for asymptomatic DCIS. [ABSTRACT FROM AUTHOR]

The incidence rate of breast fibroadenomas is higher among female kidney transplant (KT) patients treated using cyclosporine (CsA) for immunosuppression than in the general population. As such, there is an effort to convert immunosuppression from CsA or tacrolimus to sirolimus. Our aim was to assess the reversibility of a breast fibroadenoma after conversion in a small cohort of female KT recipients. This was an open-label, single-arm study including 128 female KT recipients, with a positive finding of a breast fibroadenoma in 15. Lesions were classified according to the Breast Imaging Reporting and Data System (BIRADS). Among these 15, a total of 7 converted from tacrolimus to sirolimus and 8 converted from CsA. We measured the change in BIRADS category and hormone and cytokine levels from baseline to 12 months after conversion. The primary outcome was progression or reversal of existing fibroadenomas at 12 months after conversion. Secondary outcomes were differences in hormone and cytokine levels. Conversion from CsA or tacrolimus to sirolimus had no significant effect on the BIRADS classification. However, conversion to sirolimus did produce a significant decrease in the level of transforming growth factor β cytokine, this level being closely associated with fibroadenomas. Conversion from a calcineurin inhibitor to sirolimus can block the progression of fibroadenomas. Further research is needed to confirm our results. • Cyclosporine after kidney transplant increases the risk of breast fibroadenoma. • Conversion from cyclosporine or tacrolimus to sirolimus can lower this risk. • Among 15 women with breast fibroadenoma, conversion did not change the Breast Imaging Reporting and Data System class. • Conversion did lower transforming growth factor βl level, which is associated with breast fibroadenoma. [ABSTRACT FROM AUTHOR]

Objective: The purpose if this study was to describe the follow-up ultrasound (US) findings of percutaneous US-guided laser ablation for small benign breast lesions, and to define the procedural success. Background data: There are limited reports on follow-up imaging after laser ablation, because the ablated tumors were generally resected by surgery. Materials and methods: US-guided laser ablation was performed on 19 benign breast lesions in 10 patients. The median size of the lesions was 7.8 mm (range, 3-11 mm). The lesions were diagnosed as image-pathology concordant benign by US-guided core biopsy. After ablation, we performed follow-up US with an interval of 3-12 months; the median follow-up period was 32 months (range, 30-36 months). We evaluated the morphologic and size changes by using serial follow-up US. Results: US findings of the 13 cases followed within the first 6 months after ablation showed hypoechoic lesions having inner hyperechogenicity with or without a hypoechoic center. The size of these lesions during this period increased in comparison with the initial lesion size, and then decreased or became stable on serial follow-up US. On the last follow-up US examination, in 14 cases (73.6%), the lesions were invisible, 4 cases (21.1%) showed scar changes, and 1 case (5.3%) showed a hypoechoic remnant lesion. Conclusions: US findings within 6 months after laser ablation showed mixed echoic lesions with increased size, followed by gradual decrease in size. Therefore, the success of laser ablation may be evaluated by checking the changes in the lesions by comparing the size of the lesions with the first follow-up US within 6 months and not the original tumor. [ABSTRACT FROM AUTHOR]

We present the case of 51-year-old woman with a primary breast leiomyosarcoma. On the postoperative screening after breast cancer surgery, a 4-mm-sized mass assessed as BI-RADS category 3 was detected on her contralateral breast. We recommended follow-up and observation. However, the lesion increased in size and became palpable. It was excised and diagnosed as primary leiomyosarcoma. We review the literature on imaging findings and management of breast leiomyosarcoma, and discuss the differential diagnoses in breast imaging. [ABSTRACT FROM AUTHOR]