Objectives: To establish a newly-designed scoring system for breast imaging-reporting and data system (BI-RADS) 4 and 5 breast lesions only visible on MRI, and to examine their clinical pathway of biopsy. Methods: The BI-RADS 4 and 5 breast lesions only visible on MRI but not suspected on mammograms or ultrasound between June 2007 and December 2021 at Beijing Hospital were evaluated retrospectively. A total of 209 lesions from 184 patients were finally included. All patients were female, aged (50±11) years (range: 27 to 76 years). All lesions were confirmed by pathology and divided into malignancy and non-malignancy. The lesions were divided into mass and non-mass type using BI-RADS. The receiver operator characteristic (ROC) curve was used to evaluate the diagnostic performance of the new scoring system. Four types of pathology-obtaining pathway were used: biopsy guided by second-look ultrasound, local excision guided by lesion position information on MRI, intraductal lesion excision guided by methylene blue stain and mastectomy. The data between mass and non-mass lesions were compared by Mann-Whitney U test, χ² test or Fisher exact test,respectively. Results: There were 124 malignant and 85 non-malignant lesions, while 100 mass and 109 non-mass lessions. The sizes between mass and non-mass lesions showed significant difference(M(IQR)) (7.0 (3.0) mm vs. 25.0 (25.0) mm, U=568.000, P<0.01) and their BI-RADS diagnostic accuracy had no significant difference (53.0% (53/100) vs. 65.1% (71/109), χ²=3.184, P=0.074). The areas under ROC curve of the new scoring system for evaluating mass and non-mass were 0.841 and 0.802, respectively. When taking Score 3 as threshold, it can potentially avoid 14.0% (14/100) and 4.6% (5/109) of biopsies in mass and non-mass, respectively. As to pathway of obtaining pathology, second-look ultrasound succeeded more easily in mass than non-mass (41.0% (41/100) vs.26.6% (29/109), χ²=4.851, P=0.028). More MRI-guided local excisions were performed in non-mass than mass (52.3% (57/109) vs. 34.0% (34/100), χ²=7.100, P=0.008). Conclusions: For suspicious breast lesions detected by MRI but not suspected on X-ray or ultrasound, the new scoring system can further increase diagnostic accuracy. The second-look ultrasound plays an important role for obtaining pathology, especially for mass-type lesion.
Humans ; Female ; Male ; Retrospective Studies ; Breast Neoplasms/diagnostic imaging* ; Mastectomy ; Radiography ; Magnetic Resonance Imaging

Breast cancer has now become the leading cancer in women. The development of breast ultrasound artificial intelligence (AI) diagnostic technology is conducive to promoting the precise diagnosis and treatment of breast cancer and alleviating the heavy medical burden due to the unbalanced regional development in China. In recent years, on the basis of improving diagnostic efficiency, AI technology has been continuously combined with various clinical application scenarios, thereby providing more comprehensive and reliable evidence-based suggestions for clinical decision-making. Although AI diagnostic technologies based on conventional breast ultrasound gray-scale images and cutting-edge technologies such as three-dimensional (3D) imaging and elastography have been developed to some extent, there are still technical pain points, diffusion difficulties and ethical dilemmas in the development of AI diagnostic technologies for breast ultrasound.
Artificial Intelligence ; Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Elasticity Imaging Techniques/methods* ; Female ; Humans ; Ultrasonography, Mammary/methods*

OBJECTIVE: To evaluate the value of ultrasound S-Detect in the diagnosis of breast masses. METHODS: A total of 85 breast masses in 62 female patients were diagnosed by S-Detect technique and conventional ultrasound. The diagnostic efficacy of conventional ultrasound and S-Detect technique was analyzed and compared with postoperative pathological results as the gold standard. RESULTS: When operated by junior physicians, the diagnostic efficacy of conventional ultrasound was significantly lower than that of S-Detect technique (P < 0.05), but this difference was not observed in moderately experienced and senior physicians (P>0.05). S-Detect technique was positively correlated with the diagnostic results of senior physicians (r=0.97). Using S-Detect technique, the diagnostic efficacy did not differ significantly between the long axis section and its vertical section (P>0.05). Routine ultrasound showed a better diagnostic efficacy than S-Detect for breast masses with a diameter below 20 mm (P < 0.05), but for larger breast masses, its diagnostic efficacy was significantly lower than that of SDetect (P < 0.05). CONCLUSION S-Detect can be used in differential diagnosis of benign and malignant breast masses, and its diagnostic efficiency can be comparable with that of BI-RADS classification for moderately experienced and senior physicians, but its diagnostic efficacy can be low for breast masses less than 20 mm in diameter.
Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Diagnosis, Differential ; Female ; Humans ; Sensitivity and Specificity ; Ultrasonics ; Ultrasonography ; Ultrasonography, Mammary/methods*

OBJECTIVES: Accurate breast lesion surface localization can guarantee accurate biopsy and local treatment. But there is no guideline to regular equipment and methods for the localization of breast lesions. The conventional non-invasive localization method is marker-based localization. The advantages of this method are simple and efficient. The disadvantages are that markers disappear easily under coupling agents; the positioning length of markers cannot last long on skin; and healthcare associated infection due to many patients using the same marker pen is potentially unavoidable. Breast lesion sticker (called sticker for short) is a new-type localization medical instrument in 2020. Our study aims to explore the clinical value of a new lesion stickers in breast lesion surface localization via comparison of the sticker and marker pen localization methods. METHODS: This was a prospective cohort study. It was conducted in 67 patients who needed breast lesion surface localization before biopsy. The patients were randomly assigned into 2 groups. One group of patients used marker pen to mark breast lesion surface location by ultrasonography. The other group of patients used stickers. Patients labeled with markers on skin were swabbed agents before marking. Then the markers were checked by ultrasound scan. If the surface positions of breast lesion were not correct, the above procedure was repeated. In the sticker group, the stickers were released synchronously after the lesions were detected by ultrasound scan. Then locations were checked via scanning hole. If the surface positions of breast lesion were not correct, the above procedure was repeated. The accuracy of positioning, the length of positioning time and satisfaction of patients between the 2 groups were compared. The length of positioning time was calculated from the time when ultrasound detected the lesion to the time when the surface position of breast lesion was confirmed. The total score of patients' satisfaction was 5 points according to Service Quality Evaluation of SERVQUAL Scale, including sonographers' service attitude and their technical proficiency, other medical staffs' service attitude and their technical proficiency, hospital service procedures, positioning comfort, and positioning effects. RESULTS: All 67 patients were females, aged 18-66 (39.73±13.10). There were 35 patients in the marker pen group and 32 patients in the sticker group. The time length of group used marker pen to localization was 22-88 (52.20±2.90) s, and the sticker group was 3-15 (9.22±0.58) s in length. The length of positioning time for the stickers was significantly shorter than that of the marker (P<0.01). Both methods were accurate in the surface localization of lesions before operation. The total scores of patients' satisfaction was 4-5 (4.92±0.02) in the stickers group, and 1-5 (3.35±0.10) in the marker pen group. The patients' satisfaction scores with the sticker were significantly higher than those with the marker pen (P<0.01). The length of positioning time and patients' satisfication scores for sonographer with 20 years' working experience were shorter and higher than those of sonographer with 10 years' working experience, respectively (both P<0.05). CONCLUSIONS The new breast lesion positioning stickers have more advantages than the marker pen in localization efficiency. It could reduce the workload of medical workers and increase patients' satisfaction to some extent. The stickers can be used not only in the breast lesions surface localization, but also in the skin location of pleural effusion and ascites, the skin location of surface masses, the skin location of thyroid nodule, and many other clinical marker areas, to further expand the scope of clinical application and value of the stickers.
Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Female ; Humans ; Male ; Prospective Studies ; Skin

OBJECTIVE: To investigate the value of quantitative synthetic magnetic resonance imaging (SyMRI) in distinguishing between benign and malignant breast lesions. METHODS: We retrospectively collected data of preoperative conventional MRI and multi-dynamic multi-echo sequences from 95 patients with breast lesions showing mass-type enhancement on DCE-MRI, including 27 patients with benign lesions and 68 with malignant lesions. The MRI features of the lesions (shape, margin, internal enhancement pattern, time-signal intensity curve, and T2WI signal) were analyzed, and for each lesion, SyMRI-generated quantitative parameters including T1 and T2 relaxation time and proton density (PD) were measured before and after enhancement and recorded as T1p, T2p, PDp and T1e, T2e, and PDe, respectively. The relative change rate of each parameter was calculated. Logistic regression and all-subset regression analyses were performed for variable selection to construct diagnostic models of the breast lesions, and receiver-operating characteristic (ROC) analysis was used to assess the performance of each model for differentiation of benign and malignant lesions. RESULTS: There were significant differences in the MRI features between benign and malignant lesions (P < 0.05). All the SyMRI-generated quantitative parameters, with the exception of T2e and Pdp, showed significant differences between benign and malignant lesions (P < 0.05). Among the constructed diagnostic models, the model based on all the DCE-MRI features combined with SyMRI parameters T2p and T1e (DCE-MRI+T2p+T1e) showed the best performance in the differential diagnosis malignant breast masses with an AUC of 0.995 (95% CI: 0.983-1.000). CONCLUSION Quantitative SyMRI can be used for differential diagnosis of benign and malignant breast lesions.
Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Contrast Media ; Diagnosis, Differential ; Female ; Humans ; Magnetic Resonance Imaging/methods* ; ROC Curve ; Retrospective Studies

Bone and Bones/diagnostic imaging* ; Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Female ; Humans

Triple-negative breast cancer is a complex type of breast cancer,the most common malignant tumor in women.Since the early image features of triple-negative breast cancer appear benign tumor with rapid growth,this cancer has progressed into the middle and late stages once diagnosed,which leads to high mortality.Therefore,the diagnosis of triple-negative breast cancer has always been a clinical difficulty.This article summarizes the role of ultrasound in the diagnosis and treatment of triple-negative breast cancer.The extracted multi-mode ultrasound features will facilitate the early detection of this cancer and improve the prognosis of these patients.
Breast Neoplasms/diagnostic imaging* ; Diagnosis, Differential ; Female ; Humans ; Triple Negative Breast Neoplasms/diagnostic imaging* ; Ultrasonography ; Ultrasonography, Mammary

BACKGROUND: The current deep learning diagnosis of breast masses is mainly reflected by the diagnosis of benign and malignant lesions. In China, breast masses are divided into four categories according to the treatment method: inflammatory masses, adenosis, benign tumors, and malignant tumors. These categorizations are important for guiding clinical treatment. In this study, we aimed to develop a convolutional neural network (CNN) for classification of these four breast mass types using ultrasound (US) images. METHODS: Taking breast biopsy or pathological examinations as the reference standard, CNNs were used to establish models for the four-way classification of 3623 breast cancer patients from 13 centers. The patients were randomly divided into training and test groups (n = 1810 vs. n = 1813). Separate models were created for two-dimensional (2D) images only, 2D and color Doppler flow imaging (2D-CDFI), and 2D-CDFI and pulsed wave Doppler (2D-CDFI-PW) images. The performance of these three models was compared using sensitivity, specificity, area under receiver operating characteristic curve (AUC), positive (PPV) and negative predictive values (NPV), positive (LR+) and negative likelihood ratios (LR-), and the performance of the 2D model was further compared between masses of different sizes with above statistical indicators, between images from different hospitals with AUC, and with the performance of 37 radiologists. RESULTS: The accuracies of the 2D, 2D-CDFI, and 2D-CDFI-PW models on the test set were 87.9%, 89.2%, and 88.7%, respectively. The AUCs for classification of benign tumors, malignant tumors, inflammatory masses, and adenosis were 0.90, 0.91, 0.90, and 0.89, respectively (95% confidence intervals [CIs], 0.87-0.91, 0.89-0.92, 0.87-0.91, and 0.86-0.90). The 2D-CDFI model showed better accuracy (89.2%) on the test set than the 2D (87.9%) and 2D-CDFI-PW (88.7%) models. The 2D model showed accuracy of 81.7% on breast masses ≤1 cm and 82.3% on breast masses >1 cm; there was a significant difference between the two groups (P < 0.001). The accuracy of the CNN classifications for the test set (89.2%) was significantly higher than that of all the radiologists (30%). CONCLUSIONS: The CNN may have high accuracy for classification of US images of breast masses and perform significantly better than human radiologists. TRIAL REGISTRATION Chictr.org, ChiCTR1900021375; http://www.chictr.org.cn/showproj.aspx?proj=33139.
Area Under Curve ; Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; China ; Deep Learning ; Humans ; ROC Curve ; Sensitivity and Specificity

Objective To explore the value of elastography strain ratio(SR)combined with breast ultrasound imaging reporting and data system(BI-RADS-US)in the differential diagnosis of breast nodules.Methods A total of 471 breast nodules(from 471 patients)were reclassified by SR combined with BI-RADS-US.With the pathology results as gold standard,the area under the receiver operating characteristic(ROC)curve(AUC)was employed to evaluate the diagnostic performance,and the sensitivity,specificity,and accuracy were compared between the combined method and BI-RADS-US.Results Among the 471 breast nodules,180 nodules were benign and 291 were malignant.The AUC of the combined method was statistically significantly higher than that of BI-RADS-US(0.798 vs. 0.730;Z= 2.583, P= 0.010).SR,BI-RADS-US,and the combined method for diagnosing breast nodules had the sensitivity of 86.6%,99.0%,and 96.6%,the specificity of 67.2%,47.2%,and 63.3%,and the accuracy of 79.2%,79.2%,and 83.9%,respectively.The combined method increased the specificity from 47.2%(BI-RADS-US)to 63.3%(χ
Breast/diagnostic imaging* ; Breast Neoplasms/diagnostic imaging* ; Diagnosis, Differential ; Elasticity Imaging Techniques ; Female ; Humans ; ROC Curve ; Sensitivity and Specificity ; Ultrasonography, Mammary

Both feature representation and classifier performance are important factors that determine the performance of computer-aided diagnosis (CAD) systems. In order to improve the performance of ultrasound-based CAD for breast cancers, a novel multiple empirical kernel mapping (MEKM) exclusivity regularized machine (ERM) ensemble classifier algorithm based on self-paced learning (SPL) is proposed, which simultaneously promotes the performance of both feature representation and the classifier. The proposed algorithm first generates multiple groups of features by MEKM to enhance the ability of feature representation, which also work as the kernel transform in multiple support vector machines embedded in ERM. The SPL strategy is then adopted to adaptively select samples from easy to hard so as to gradually train the ERM classifier model with improved performance. This algorithm is verified on a B-mode ultrasound dataset and an elastography ultrasound dataset, respectively. The results show that the classification accuracy, sensitivity and specificity on B-mode ultrasound are (86.36±6.45)%, (88.15±7.12)%, and (84.52±9.38)%, respectively, and the classification accuracy, sensitivity and specificity on elastography ultrasound are (85.97±3.75)%, (85.93±6.09)%, and (86.03±5.88)%, respectively. It indicates that the proposed algorithm can effectively improve the performance of ultrasound-based CAD for breast cancers with the potential for application.
Algorithms ; Breast Neoplasms/diagnostic imaging* ; Computers ; Diagnosis, Computer-Assisted ; Humans ; Support Vector Machine ; Ultrasonography