Breast Imaging Reporting And Data System Features Research Articles

Characterization of breast lumps locations using high-frequency ultrasound: Correlation between Doppler vascularity and BI-RADS

BackgroundA breast lump is a major concern, and it is critical for every clinician to diagnose and manage patients with breast lumps. High-frequency ultrasound (U/S) is the first imaging technique for evaluating a breast lump and classifying it as benign or malignant based on many ultrasound criteria; Breast imaging-reporting and data system (BI-RADS) and Doppler vascularity also contribute important information for identification. The main aim of the current study was to characterize breast lesions using high-frequency ultrasound with grayscale and Doppler as well as correlate Doppler vascularity and BI-RADS features. Materials and methodsA retrospective study was conducted in Madinah–Saudi Arabia at Al-Madinah General Hospital from November 2021 to March 2022. The sampling included 124 females' breast ultrasound reports and was analyzed (by radiologist) to demonstrate the features of breast masses shown through ultrasound. ResultsMost study samples were aged 18–38 years (66.1%). The most common complaints were lump and pain (31.45% and 9.68%, respectively). About (31.3%) and most of lumps located in upper outer quadrant UOQ, (42.2%), hypoechoic and anechoic (71.1%, 34.9%, respectively), well-defined (45.2%), oval (68.6%) and a vascular (85.5%). Most lumps had a BIRADS feature of 0–3, and there was no significant correlation between female age and complaints (p-value> 0.05). ConclusionThe study concluded among the study sample most breast lesions were scattered and located in UOQ, and most of them were hypoechoic and anechoic, well-defined, and a vascular, most of them have BI-RADs characteristics 0–3, there is a significant correlation between breast lumps’ BI-RADS features and Doppler vascularity.

Multi-Task/Single-Task Joint Learning of Ultrasound BI-RADS Features.

Breast cancer is one of the most common cancers among women in the world. The breast imaging reporting and data system (BI-RADS) features effectively improve the accuracy and sensitivity of breast tumors. Based on the description of signs in BI-RADS, a quantitative scoring scheme is proposed based on ultrasound (US) data. This scheme includes feature extraction of high-level semantic, that is, an intermediate step interpreting the subsequent diagnosis. However, the scheme requires doctors to score the features of breast data, which is labor-intensive. To reduce the burden of doctors, we design a multi-task learning (MTL) framework, which can directly output the scores of different BI-RADS features from the raw US images. The MTL framework consists of a shared network that learns global features and K soft attention networks for different BI-RADS features. Thus, it enables the network not only to learn the potential correlation among different BI-RADS features, but also learn the unique specificity of each feature, which can assist each other and jointly improve the scoring accuracy. In addition, we group different BI-RADS features according to the correlation among tasks and build a multi-task/single-task joint framework. Experimental results on the US breast tumor dataset collected from 1859 patients with 4458 US images show that the proposed BI-RADS feature scoring framework achieves an average scoring accuracy of 84.91% for 11 BI-RADS features on the test dataset, which is helpful for the subsequent diagnosis of breast tumors.

Correlation of ultrasound, cytological, and histological features of 110 benign BI-RADS categories 4C and 5 nonpalpable breast lesions. The Institut Curie's experience.

The purpose of this study was to determine the pathological and ultrasound (US) features of benign nonpalpable breast lesions (NPBLs) classified as Breast Imaging Reporting and Data System (BI-RADS) category 4C or 5. Between 2003 and 2007, 849 consecutive NPBLs detected at US and classified as BI-RADS category 4C (505)or 5 (344) initially underwent US-guided fine needle aspiration (FNA) at our institution. Benign diagnoses were established according to surgical excision findings or during a minimal 6-month imaging follow-up (mean, 3.7 years [SD, 2.6 years]). US BI-RADS features were reviewed and compared retrospectively using a chi-square test for the following pathological categories: epithelial and fibrous proliferation (EFP), cystic and papillary lesion (C&P), inflammatory lesion (IL), benign tumor (BT), intramammary lymph node (ILN), intraepithelial proliferative lesion (IPL), and nonspecific morphological alteration (NMA). The performance of FNA in the diagnosis of benignity was assessed. Of 849 NPBLs, 110 (12.9%) NPBLs were benign: 88 (17.4%) were BI-RADS category 4C, and 22 (6.4%) were BI-RADS category 5. Forty-four (40%) were EFPs, 21 (19%) were C&Ps, 13 (12%) were NMAs, 11 (10%) were ILs, 11 (10%) were BTs, 8 (7%) were IPLs, and 2 (2%) were ILNs. Lesion shape, US pattern distribution, and posterior features showed statistically significant differences between these categories (P < .05): 33 (75%) EFPs exhibited posterior shadowing, 18 (86%) C&Ps were homogenous, 9 (82%) ILs were heterogeneous, 11 (100%) BTs were homogeneous, 9 (82%) BTs were oval, and 6 (75%) IPLs were irregularly shaped. Of the 110 benign NPBLs, FNA diagnosis was falsely positive in 7 (6%), suspicious in 10 (9%), and benign in 90 (82%), and 3 (3%) were inadequate for diagnosis. A diverse array of benign NPBLs can be classified as BI-RADS category 4C or 5 on US, each showing specific imaging presentations.

Study of ultrasound BI-RADS for small solid breast masses

Objective To evaluate the clinical value of ultrasound breast imaging reporting and data system （BI-RADS） in small solid breast masses with diameter 41 cm.Methods The ultrasound features of 230 solid breast masses with diameter ≤ 1 cm were described by ultrasound BI-RADS, the relationship between ultrasound features,BI-RADS final assesment and pathology were analyzed.Results Of these 230 masses,72（31.3%） were pathologically confirmed to be malignant and 158 （68.7%） to be benign. The ultrasound BI-RADS features of mass shape, margin, orientation, posterior acoustic features, and microcalcificaition were significantly different between malignant and benign masses （ P 〈 0.05）. Irregular shape, noncircumscribed, nonparallel orientation, postrior shadowing, microcalcifications were regarded as malignant ultrasound features, their positive predictive values （PPV）, sensitivity, specificity and accuracy for malignancy were 53.3% - 100% ,2.8% - 75.0% ,82.3% - 100% ,69.6% - 80.9% ,respectively. One hundred and fifty-two（66.1%）,62（27.0%）, 16（7.0%） masses were classified into grade 3,4,and 5 ,respectively.The PPV for grade 3,4 and 5 were 10.5 %, 64.5 %, 100 % respectively. Among BI-RADS grade 3 cases, 87.5 % malignant masses were intraductal carcinoma in situ and special type of invasive cancer,among pathological benign BI-RADS grade 4 masses,90.9% were hyperplasia and intraductal papilloma.Conclusions In small breast masses with diameter ≤ 1 cm, due to the sensitivity of malignant signs are not high, the overlap between signs of benign and malignant lesions, pathological type and other factors, the positive predictive value of BI-RADS grade 3 is higher than criteria of American College of Radiology, so BI-RADS classification requires further detailed study. Key words: Ultrasonography ; Breast diseases ; BI-RADS

Predicting error in detecting mammographic masses among radiology trainees using statistical models based on BI-RADS features

The purpose of this study is to explore Breast Imaging-Reporting and Data System (BI-RADS) features as predictors of individual errors made by trainees when detecting masses in mammograms. Ten radiology trainees and three expert breast imagers reviewed 100 mammograms comprised of bilateral medial lateral oblique and craniocaudal views on a research workstation. The cases consisted of normal and biopsy proven benign and malignant masses. For cases with actionable abnormalities, the experts recorded breast (density and axillary lymph nodes) and mass (shape, margin, and density) features according to the BI-RADS lexicon, as well as the abnormality location (depth and clock face). For each trainee, a user-specific multivariate model was constructed to predict the trainee's likelihood of error based on BI-RADS features. The performance of the models was assessed using area under the receive operating characteristic curves (AUC). Despite the variability in errors between different trainees, the individual models were able to predict the likelihood of error for the trainees with a mean AUC of 0.611 (range: 0.502-0.739, 95% Confidence Interval: 0.543-0.680,p < 0.002). Patterns in detection errors for mammographic masses made by radiology trainees can be modeled using BI-RADS features. These findings may have potential implications for the development of future educational materials that are personalized to individual trainees.