Breast contrast-enhanced ultrasound (CEUS)has been used in clinical practice for nearly two decades. CEUS provides real-time assessment of microvascular perfusion, offering valuable functional information complementary to conventional ultrasound. Nonetheless, the lack of standardization in relevant techniques has hindered the widespread adoption and further development of this technology. As a specialized committee, Superficial Organs and Peripheral Vessels Committee of Chinese Association of Ultrasound in Medicine and Engineering has always been committed to standardizing the clinical application of ultrasound. This expert consensus aims to standardize examination procedures, image interpretation, and reporting for diagnostic settings. It outlines key indications and provides recommendations on acquisition techniques, qualitative and quantitative analysis of enhancement patterns, and standardized lexicon for reporting.

OBJECTIVES:This study aimed to evaluate the ultrasound findings in skeletal muscle lymphoma and investigate the correlations between ultrasound findings and histopathological characteristics. METHODS:Ultrasonographic images with skeletal muscle lymphoma obtained between June 2014 and June 2024 were retrospectively reviewed. The ultrasonographic examinations comprised gray-scale (n = 35), color-Doppler (n = 35), contrast-enhanced (n = 5), and elastography imaging (n = 5). Biopsy samples underwent histopathological analysis to identify the histopathologic basis of the typical ultrasound features. Kappa statistics were used to assess the consistency between ultrasound findings and histopathological features. RESULTS:The study comprised 35 patients (mean age, 63 years ±15 [standard deviation], 17 men). In gray-scale sonography, 35 lesions (100%) appeared as hypoechoic areas in the muscle with poorly defined margins. Residual myofiber-like echoes were observed in 25 (71.4%) lesions, and a "cobblestone" appearance was noted in 18 (51.4%). In color-Doppler sonography, 10 (28.6%) patients had intact transverse vessels perpendicular to the long axis of the muscle. In contrast-enhanced sonography, 4 out of 5 (80%) lesions showed synchronous high enhancement. Four out of 5 (80%) lesions exhibited high stiffness on elastography imaging. Myofiber-like echoes (κ: 0.800, P < 0.001) and the "cobblestone" appearance (κ: 0.521, P < .017) are consistent with myofibers in pathology. CONCLUSION:Skeletal muscle lymphoma demonstrates specific features on multimodal ultrasound.

Bethesda category III thyroid nodules pose substantial management challenges, and a validated malignancy risk stratification system is urgently needed to guide clinical decision-making. This study aims to identify specific sonographic and cytological features associated with malignant Bethesda III thyroid nodules and to develop a practical model for predicting malignancy risk through an integrated analysis of multimodal data from histopathologically confirmed cases. This retrospective study used clinical data from patients with both Bethesda III cytological diagnoses and corresponding surgical pathology outcomes between July 2016 and December 2024. The sonographic, clinical, and cytological characteristics of benign and malignant nodules were systematically assessed and compared. Univariable and multivariable logistic regression analyses were conducted to evaluate the association between these features and malignancy. A prediction nomogram, incorporating sonographic and cytological features independently associated with malignancy, was developed and validated to assess its performance. In this study, a total of 187 Bethesda III thyroid nodules were analyzed, consisting of 77 benign nodules and 110 malignant nodules. Factors such as maximum diameter ≤ 1 cm, absence of smooth margins, microcalcifications, and nuclear atypia in cytology were identified as independent factors associated with malignancy. A prediction nomogram model was developed using these variables, demonstrating strong performance in distinguishing between benign and malignant nodules categorized as Bethesda III, with an area under the curve (AUC) of 0.874. In Bethesda III thyroid nodules, malignant cases were associated with suspicious sonographic characteristics, smaller size (≤ 1 cm), and cytological nuclear atypia. The nomogram developed from these predictors demonstrates potential utility in estimating malignancy risk for Bethesda III thyroid nodules.

Purpose: To evaluate the ability of computer-aided diagnosis (CAD) system (S-Detect) to identify malignancy in ultrasound (US) -detected BI-RADS 3 breast lesions. Materials and methods: 148 patients with 148 breast lesions categorized as BI-RADS 3 were included in the study between January 2021 and September 2022. The malignancy rate, accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) were calculated. Results: In this study, 143 breast lesions were found to be benign, and 5 breast lesions were malignant (malignancy rate, 3.4 %, 95 % confidence interval (CI): 0.5–6.3). The malignancy rate rose significantly to 18.2 % (4/22, 95 % CI: 2.1–34.3) in the high-risk group with a “possibly malignant” CAD result (p = 0.017). With a “possibly benign” CAD result, the malignancy rate decreased to 0.8 % (1/126, 95 % CI: 0–2.2) in the low-risk group (p = 0.297). The AUC, sensitivity, specificity, accuracy, PPV, and NPV of the CAD system in BI-RADS 3 breast lesions were 0.837 (95 % CI: 77.7–89.6), 80.0 % (95 % CI: 73.6–86.4), 87.4 % (95 % CI: 82.0–92.7), 87.2 % (95 % CI: 81.8–92.6), 18.2 % (95 % CI: 2.1–34.3) and 99.2 % (95 % CI: 97.8–100.0), respectively. Conclusions: CAD system (S-Detect) enables radiologists to distinguish a high-risk group and a low-risk group among US-detected BI-RADS 3 breast lesions, so that patients in the low-risk group can receive follow-up without anxiety, while those in the high-risk group with a significantly increased malignancy rate should actively receive biopsy to avoid delayed diagnosis of breast cancer.

A gastric GIST which causes gastroduodenal intussusception is rare. A Pubmed search only identified 21 published cases of gastroduodenal intussusception due to gastric GIST. Only 2 of them mentioned ultrasound without further analysis. Here, we report a case of gastroduodenal intussusception due to a gastric GIST with multiple imaging especially ultrasound. A 72-year-old Chinese woman was admitted to hospital because of epigastric pain, black stool lasting and occasional vomiting for 2 months. She underwent abdominal ultrasound, endoscopy, and contrast enhanced CT in turn. Abdominal ultrasound revealed a hypoechoic, medium-sized lesion beside pancreatic head. Endoscopy showed a submucosal lesion of gastric fundus overlapping into duodenum. The lesion manifested slight enhancement in the arterial phase on enhanced CT scans. The patient underwent laparoscopic exploration and partial gastrectomy. The histological examination revealed a low-risk gastric GIST of spindle-shaped cell type.

ObjectivesThis study aims to identify distinct ultrasound (US) characteristics for distinguishing follicular thyroid carcinoma (FTC) from follicular thyroid adenoma (FTA), and construct a user‐friendly preoperative risk stratification model for thyroid follicular neoplasms.MethodsIn this retrospective study, patients diagnosed with pathologically confirmed FTA or FTC and undergoing US examinations between July 2017 and June 2021 were designated as the training cohort, and those from July 2021 to June 2023 were enrolled as the external validation set. We systematically assessed and compared the sonographic and clinical characteristics of FTC and FTA. Univariable and multivariable logistic regression analyses were used to assess the association of US features with FTC in the training set. A prediction nomogram model, incorporating US features independently associated with FTC, was developed and validated externally to assess its performance.ResultsA total of 645 patients (FTA/FTC = 530/115) were included in the training set, while 197 patients (FTA/FTC = 165/32) constituted the validation set. In the training set, solid composition, hypo‐echogenicity, irregular margin, calcification, protrusion sign, trabecular formation, absent or thick halo, and mainly central hypervascularity were identified as independent factors associated with FTC. The prediction nomogram model constructed using these variables showed good performance in differentiating FTC from FTA with an area under the curve of 0.948 in the training set and 0.915 in the validation set.ConclusionsThe preoperative nomogram model constructed based on US features serves as an effective tool for the risk stratification of thyroid follicular neoplasms.

BACKGROUND:Segmenting liver vessels from contrast-enhanced computed tomography images is essential for diagnosing liver diseases, planning surgeries and delivering radiotherapy. Nevertheless, identifying vessels is a challenging task due to the tiny cross-sectional areas occupied by vessels, which has posed great challenges for vessel segmentation, such as limited features to be learned and difficult to construct high-quality as well as large-volume data.METHODS:We present an approach that only requires a few labeled vessels but delivers significantly improved results. Our model starts with vessel enhancement by fading out liver intensity and generates candidate vessels by a classifier fed with a large number of image filters. Afterwards, the initial segmentation is refined using Markov random fields.RESULTS:In experiments on the well-known dataset 3D-IRCADb, the averaged Dice coefficient is lifted to 0.63, and the mean sensitivity is increased to 0.71. These results are significantly better than those obtained from existing machine-learning approaches and comparable to those generated from deep-learning models.CONCLUSION:Sophisticated integration of a large number of filters is able to pinpoint effective features from liver images that are sufficient to distinguish vessels from other liver tissues under a scarcity of large-volume labeled data. The study can shed light on medical image segmentation, especially for those without sufficient data.

To establish an automated, multitask, MRI-based deep learning system for the detailed evaluation of supraspinatus tendon (SST) injuries. According to arthroscopy findings, 3087 patients were divided into normal, degenerative, and tear groups (groups 0–2). Group 2 was further divided into bursal-side, articular-side, intratendinous, and full-thickness tear groups (groups 2.1–2.4), and external validation was performed with 573 patients. Visual geometry group network 16 (VGG16) was used for preliminary image screening. Then, the rotator cuff multitask learning (RC-MTL) model performed multitask classification (classifiers 1–4). A multistage decision model produced the final output. Model performance was evaluated by receiver operating characteristic (ROC) curve analysis and calculation of related parameters. McNemar’s test was used to compare the differences in the diagnostic effects between radiologists and the model. The intraclass correlation coefficient (ICC) was used to assess the radiologists’ reliability. p < 0.05 indicated statistical significance. In the in-group dataset, the area under the ROC curve (AUC) of VGG16 was 0.92, and the average AUCs of RC-MTL classifiers 1–4 were 0.99, 0.98, 0.97, and 0.97, respectively. The average AUC of the automated multitask deep learning system for groups 0–2.4 was 0.98 and 0.97 in the in-group and out-group datasets, respectively. The ICCs of the radiologists were 0.97–0.99. The automated multitask deep learning system outperformed the radiologists in classifying groups 0–2.4 in both the in-group and out-group datasets (p < 0.001). The MRI-based automated multitask deep learning system performed well in diagnosing SST injuries and is comparable to experienced radiologists. Our study established an automated multitask deep learning system to evaluate supraspinatus tendon (SST) injuries and further determine the location of SST tears. The model can potentially improve radiologists’ diagnostic efficiency, reduce diagnostic variability, and accurately assess SST injuries. • A detailed classification of supraspinatus tendon tears can help clinical decision-making. • Deep learning enables the detailed classification of supraspinatus tendon injuries. • The proposed automated multitask deep learning system is comparable to radiologists.

To evaluate the value of ultrasound-guided vacuum-assisted excision (US-guided VAE) in the treatment of intraductal papillomas, including intraductal papillomas with atypical ductal hyperplasia (ADH), and to evaluate the lesion characteristic features affecting the local recurrence rate. Between August 2011 and December 2020, 91 lesions of 91 patients underwent US-guided VAE and were diagnosed with intraductal papilloma with or without ADH. The recurrence rate of intraductal papilloma was evaluated on follow-up US. The lesion characteristic features were analyzed to identify the factors affecting the local recurrence rate. The local recurrence rate of intraductal papillomas removed by US-guided VAE was 7.7% (7/91), with the follow-up duration 12–92 months (37.4 ± 23.9 months). Of the 91 patients, five cases diagnosed as intraductal papilloma with ADH did not recur, with the follow-up time 12–47 months (26.4 ± 14.4 months). There were no malignant transformation in all 91 cases during the follow-up period. All 7 patients recurred 7–58 months (22.8 ± 19.2 months) after US-guided VAE. There were no significant differences between the non-recurrence and recurrence groups in terms of age, side, distance from nipple, lesion size, BI-RADS category, with ADH, or history of excision (p > 0.05). US-guided VAE is an effective method for the treatment of intraductal papilloma, including intraductal papilloma with ADH. It avoids invasive surgical excision, but regular follow-up is recommended to prevent recurrence or new onset due to multifocality. Any suspicious lesions during the follow-up should be actively treated.

Background Most of superficial soft-tissue masses are benign tumors, and very few are malignant tumors. However, persistent growth, of both benign and malignant tumors, can be painful and even life-threatening. It is necessary to improve the differential diagnosis performance for superficial soft-tissue masses by using deep learning models. This study aimed to propose a new ultrasonic deep learning model (DLM) system for the differential diagnosis of superficial soft-tissue masses. Methods Between January 2015 and December 2022, data for 1615 patients with superficial soft-tissue masses were retrospectively collected. Two experienced radiologists (radiologists 1 and 2 with 8 and 30 years’ experience, respectively) analyzed the ultrasound images of each superficial soft-tissue mass and made a diagnosis of malignant mass or one of the five most common benign masses. After referring to the DLM results, they re-evaluated the diagnoses. The diagnostic performance and concerns of the radiologists were analyzed before and after referring to the results of the DLM results. Results In the validation cohort, DLM-1 was trained to distinguish between benign and malignant masses, with an AUC of 0.992 (95% CI: 0.980, 1.0) and an ACC of 0.987 (95% CI: 0.968, 1.0). DLM-2 was trained to classify the five most common benign masses (lipomyoma, hemangioma, neurinoma, epidermal cyst, and calcifying epithelioma) with AUCs of 0.986, 0.993, 0.944, 0.973, and 0.903, respectively. In addition, under the condition of the DLM-assisted diagnosis, the radiologists greatly improved their accuracy of differential diagnosis between benign and malignant tumors. Conclusions The proposed DLM system has high clinical application value in the differential diagnosis of superficial soft-tissue masses.

To evaluate the relevant factors associated with malignancy in Breast Imaging Reporting and Data System (BI-RADS) 4A and to determine whether it was possible to establish a safe follow-up guideline for lower-risk 4A lesions. In this retrospective study, patients categorized as BI-RADS 4A on ultrasound who underwent ultrasound-guided biopsy or/and surgery between June 2014 and April 2020 was analyzed. Classification-tree method and cox regression analysis were used to explore the possible correlation factors of malignancy. Among 9965 patients enrolled, 1211 (mean age, 44.3 ± 13.5 years; range, 18–91 years) patients categorized as BI-RADS 4A were eligible. The result of cox regression analysis revealed the malignant rate was only associated with patient age (hazard ratio (HR) = 1.038, p < 0.001, 95

To investigate the potential applicability of AI-assisted compressed sensing (ACS) in knee MRI to enhance and optimize the scanning process. Volunteers and patients with sports-related injuries underwent prospective MRI scans with a range of acceleration techniques. The volunteers were subjected to varied ACS acceleration levels to ascertain the most effective level. Patients underwent scans at the determined optimal 3D-ACS acceleration level, and 3D compressed sensing (CS) and 2D parallel acquisition technology (PAT) scans were performed. The resultant 3D-ACS images underwent 3.5 mm/2.0 mm multiplanar reconstruction (MPR). Experienced radiologists evaluated and compared the quality of images obtained by 3D-ACS-MRI and 3D-CS-MRI, 3.5 mm/2.0 mm MPR and 2D-PAT-MRI, diagnosed diseases, and compared the results with the arthroscopic findings. The diagnostic agreement was evaluated using Cohen’s kappa correlation coefficient, and both absolute and relative evaluation methods were utilized for objective assessment. The study involved 15 volunteers and 53 patients. An acceleration factor of 10.69 × was identified as optimal. The quality evaluation showed that 3D-ACS provided poorer bone structure visualization, and improved cartilage visualization and less satisfactory axial images with 3.5 mm/2.0 mm MPR than 2D-PAT. In terms of objective evaluation, the relative evaluation yielded satisfactory results across different groups, while the absolute evaluation revealed significant variances in most features. Nevertheless, high levels of diagnostic agreement (κ: 0.81–0.94) and accuracy (0.83–0.98) were observed across all diagnoses. ACS technology presents significant potential as a replacement for traditional CS in 3D-MRI knee scans, allowing thinner MPRs and markedly faster scans without sacrificing diagnostic accuracy. 3D-ACS-MRI of the knee can be completed in the 160 s with good diagnostic consistency and image quality. 3D-MRI-MPR can replace 2D-MRI and reconstruct images with thinner slices, which helps to optimize the current MRI examination process and shorten scanning time. • AI-assisted compressed sensing technology can reduce knee MRI scan time by over 50 • 3D AI-assisted compressed sensing MRI and related multiplanar reconstruction can replace traditional accelerated MRI and yield thinner 2D multiplanar reconstructions. • Successful application of 3D AI-assisted compressed sensing MRI can help optimize the current knee MRI process.

Background: Diagnosing anterior talofibular ligament (ATFL) injuries differs among radiologists. Further assessment of ATFL tears is valuable for clinical decision-making.Purpose: To establish a deep learning method for classifying ATFL injuries based on magnetic resonance imaging (MRI).Study Type: Retrospective.Population: One thousand seventy-three patients from a single center with ankle MRI within 1 month of reference stan-dard arthroscopy (in-group dataset), were divided into training, validation, and test sets in a ratio of 8:1:1. Additionally, 167 patients from another center were used as an independent out-group dataset.Field Strength/Sequence: Fat-saturation proton density-weighted fast spin-echo sequence at 1.5/3.0 T.Assessment: Patients were divided into normal, strain and degeneration, partial tear and complete tear groups (groups 0-3). The complete tear group was divided into five sub-groups by location and the potential avulsion fracture (groups 3.1-3.5). All images were input into AlexNet, VGG11, Small-Sample-Attention Net (SSA-Net), and SSA-Net + Weight Loss for classification. The results were compared with four radiologists with 5-30 years of experience.Statistical Tests: Model performance was evaluated by the receiver operating characteristic (ROC) curve, the area under the ROC curve (AUC), and so on. McNemar's test was used to compare performance among the different models, and between the radiologists and models. The intraclass correlation coefficient (ICC) was used to assess the reliability of the radiologists. P < 0.05 was considered statistically significant.Results: The average AUC of AlexNet, VGG11, SAA-Net, and SSA-Net + Weight Loss was 0.95, 0.99, 0.99, 0.99 in groups 0-3 and 0.96, 0.99, 0.99, 0.99 in groups 3.1-3.5. The effect of SSA-Net + Weight Loss was similar to SSA-Net but better than AlexNet and VGG11. In the out-group test set, the AUC of SSA-Net + Weight Loss ranged from 0.89 to 0.99. The ICC of radiologists was 0.97-1.00. The effect of SSA-Net + Weight Loss was better than each radiologist in the in-group and out-group test sets.Data Conclusion: Deep learning has potential to be used for classifying ATFL injuries. SSA-Net + Weight Loss has a better diagnostic effect than radiologists with different experience levels.

Background:The classification of calcaneofibular ligament (CFL) injuries on magnetic resonance imaging (MRI) is time-consuming and subject to substantial interreader variability. This study explores the feasibility of classifying CFL injuries using deep learning methods by comparing them with the classifications of musculoskeletal (MSK) radiologists and further examines image cropping screening and calibration methods.Methods:The imaging data of 1,074 patients who underwent ankle arthroscopy and MRI examinations in our hospital were retrospectively analyzed. According to the arthroscopic findings, patients were divided into normal (class 0, n=475); degeneration, strain, and partial tear (class 1, n=217); and complete tear (class 2, n=382) groups. All patients were divided into training, validation, and test sets at a ratio of 8:1:1. After preprocessing, the images were cropped using Mask region-based convolutional neural network (R-CNN), followed by the application of an attention algorithm for image screening and calibration and the implementation of LeNet-5 for CFL injury classification. The diagnostic effects of the axial, coronal, and combined models were compared, and the best method was selected for outgroup validation. The diagnostic results of the models in the intragroup and outgroup test sets were compared with those results of 4 MSK radiologists of different seniorities.Results:The mean average precision (mAP) of the Mask R-CNN using the attention algorithm for the left and right image cropping of axial and coronal sequences was 0.90-0.96. The accuracy of LeNet-5 for classifying classes 0-2 was 0.92, 0.93, and 0.92, respectively, for the axial sequences and 0.89, 0.92, and 0.90, respectively, for the coronal sequences. After sequence combination, the classification accuracy for classes 0-2 was 0.95, 0.97, and 0.96, respectively. The mean accuracies of the 4 MSK radiologists in classifying the intragroup test set as classes 0-2 were 0.94, 0.91, 0.86, and 0.85, all of which were significantly different from the model. The mean accuracies of the MSK radiologists in classifying the outgroup test set as classes 0-2 were 0.92, 0.91, 0.87, and 0.85, with the 2 senior MSK radiologists demonstrating similar diagnostic performance to the model and the junior MSK radiologists demonstrating worse accuracy.Conclusions:Deep learning can be used to classify CFL injuries at similar levels to those of MSK radiologists. Adding an attention algorithm after cropping is helpful for accurately cropping CFL images.

BACKGROUND. Computer-aided diagnosis (CAD) systems for breast ultrasound interpretation have been primarily evaluated at tertiary and/or urban medical centers by radiologists with breast ultrasound expertise. OBJECTIVE. The purpose of this study was to evaluate the usefulness of deep learning-based CAD software on the diagnostic performance of radiologists without breast ultrasound expertise at secondary or rural hospitals in the differentiation of benign and malignant breast lesions measuring up to 2.0 cm on ultrasound. METHODS. This prospective study included patients scheduled to undergo biopsy or surgical resection at any of eight participating secondary or rural hospitals in China of a breast lesion classified as BI-RADS category 3-5 on prior breast ultrasound from November 2021 to September 2022. Patients underwent an additional investigational breast ultrasound, performed and interpreted by a radiologist without breast ultrasound expertise (hybrid body/breast radiologists, either who lacked breast imaging subspecialty training or for whom the number of breast ultrasounds performed annually accounted for less than 10% of all ultrasounds performed annually by the radiologist), who assigned a BI-RADS category. CAD results were used to upgrade reader-assigned BI-RADS category 3 lesions to category 4A and to downgrade reader-assigned BI-RADS category 4A lesions to category 3. Histologic results of biopsy or resection served as the reference standard. RESULTS. The study included 313 patients (mean age, 47.0 +/- 14.0 years) with 313 breast lesions (102 malignant, 211 benign). Of BI-RADS category 3 lesions, 6.0% (6/100) were upgraded by CAD to category 4A, of which 16.7% (1/6) were malignant. Of category 4A lesions, 79.1% (87/110) were downgraded by CAD to category 3, of which 4.6% (4/87) were malignant. Diagnostic performance was significantly better after application of CAD, in comparison with before application of CAD, in terms of accuracy (86.6% vs 62.6%, p < .001), specificity (82.9% vs 46.0%, p < .001), and PPV (72.7% vs 46.5%, p < .001) but not significantly different in terms of sensitivity (94.1% vs 97.1%, p = .38) or NPV (96.7% vs 97.0%, p > .99). CONCLUSION. CAD significantly improved radiologists' diagnostic performance, showing particular potential to reduce the frequency of benign breast biopsies. CLINICAL IMPACT. The findings indicate the ability of CAD to improve patient care in settings with incomplete access to breast imaging expertise.

Purpose The aim of this study was to analyze the role of ultrasound-guided vacuum-assisted excision (US-guided VAE) in the treatment of high-risk breast lesions and to evaluate the clinical and US features of the patients associated with recurrence or development of malignancy. Materials and methods Between April 2010 and September 2021, 73 lesions of 73 patients underwent US-guided VAE and were diagnosed with high-risk breast lesions. The incidence of recurrence or development of malignancy for high-risk breast lesions was evaluated at follow-up period. The clinical and US features of the patients were analyzed to identify the factors affecting the recurrence or development of malignancy rate. Results Only benign phyllodes tumors on US-guided VAE showed recurrences, while other high-risk breast lesions that were atypical ductal hyperplasia (ADH), lobular neoplasia (atypical lobular hyperplasia/lobular carcinoma in situ), radial scar, and flat epithelial atypia did not show recurrences or malignant transformation. The recurrence rate of the benign phyllodes tumor was 20.8% (5/24) in a mean follow-up period of 34.3 months. The recurrence rate of benign phyllodes tumor with distance from nipple of less than 1 cm was significantly higher than that of lesions with distance from nipple of more than 1 cm (75% vs. 10%, p < 0.05). Conclusions Benign phyllodes tumors without concurrent breast cancer could be safely followed up instead of surgical excision after US-guided VAE when the lesions were classified as BI-RADS 3 or 4A by US.

Purpose MR arthrography (MRA) is the most accurate method for preoperatively diagnosing superior labrum anterior–posterior (SLAP) lesions, but diagnostic results can vary considerably due to factors such as experience. In this study, deep learning was used to facilitate the preliminary identification of SLAP lesions and compared with radiologists of different seniority. Methods MRA data from 636 patients were retrospectively collected, and all patients were classified as having/not having SLAP lesions according to shoulder arthroscopy. The SLAP-Net model was built and tested on 514 patients (dataset 1) and independently tested on data from two other MRI devices (122 patients, dataset 2). Manual diagnosis was performed by three radiologists with different seniority levels and compared with SLAP-Net outputs. Model performance was evaluated by the receiver operating characteristic (ROC) curve, area under the ROC curve (AUC), etc. McNemar’s test was used to compare performance among models and between radiologists’ models. The intraclass correlation coefficient (ICC) was used to assess the radiologists’ reliability. p < 0.05 was considered statistically significant. Results SLAP-Net had AUC = 0.98 and accuracy = 0.96 for classification in dataset 1 and AUC = 0.92 and accuracy = 0.85 in dataset 2. In dataset 1, SLAP-Net had diagnostic performance similar to that of senior radiologists ( p = 0.055) but higher than that of early- and mid-career radiologists ( p = 0.025 and 0.011). In dataset 2, SLAP-Net had similar diagnostic performance to radiologists of all three seniority levels ( p = 0.468, 0.289, and 0.495, respectively). Conclusions Deep learning can be used to identify SLAP lesions upon initial MR arthrography examination. SLAP-Net performs comparably to senior radiologists.

目的 观察节细胞神经瘤(GN)超声表现.方法 回顾性分析18例接受超声检查的GN患者,均为单发病变,观察其超声表现.结果 18个GN最大径0.9～19.0 cm、中位最大径5.6(3.4,8.8)cm,边界均清晰,对周围脏器和血管无侵犯.13个GN病灶呈纵向生长、5个呈横向生长;11个形态不规则、质地较软,沿周围组织间隙蔓延式生长,推移而不侵犯周围组织结构,7个呈类圆形膨胀性生长.18个内部均为较均匀的实性低回声,8个瘤体内可见散在、斑块样/蛋壳样钙化灶,5个肿瘤内部和/或周边探及少许短线样血流信号.结论 GN超声表现为实性低回声包块,边界清晰,可呈膨胀性生长的类圆形或沿周围组织间隙蔓延生长的不规则形,内部可见散在钙化灶,多无明显血流信号或仅见少量血流信号.

Objectives Ultrasound tends to present very high sensitivity but relatively low specificity and positive predictive value (PPV), which would result in unnecessary breast biopsies. The purpose of this study is to analyze the diagnostic performance of computer-aided diagnosis (CAD) (S-Detect) system in differentiating breast lesions and reducing unnecessary biopsies in non-university hospitals in less-developed regions of China. Methods The study was a prospective multicenter study from 8 hospitals. The ultrasound images, and cine, CAD analysis, and BI-RADS were recorded. The accuracy, sensitivity, specificity, PPV, negative predictive value (NPV), and area under the curve (AUC) were analyzed and compared between CAD and radiologists. The Youden Index (YI) was used to determine optimal cut-off for the number of planes to downgrade. Results A total of 491 breast lesions were included in the study. Less-experienced radiologists combined CAD was superior to less-experienced radiologists alone in AUC (0.878 vs 0.712, p < 0.001), and specificity (81.3% vs 44.6%, p < 0.001). There was no statistical difference in AUC (0.891 vs 0.878, p = 0.346), and specificity (82.3% vs 81.3%, p = 0.791) between experienced radiologists and less-experienced radiologists combined CAD. With CAD assistance, the biopsy rate of less-experienced radiologists was significantly decreased (100.0% vs 25.6%, p < 0.001), and malignant rate of biopsy was significantly increased (15.0% vs 43.9%, p < 0.001). Conclusions CAD system can be an effective auxiliary tool in differentiating breast lesions and reducing unnecessary biopsies for radiologists from non-university hospitals in less-developed regions of China.

Background The diagnosis of labral injury on MRI is time‐consuming and potential for incorrect diagnoses. Purpose To explore the feasibility of applying deep learning to diagnose and classify labral injuries with MRI. Study Type Retrospective. Population A total of 1016 patients were divided into normal (n = 168, class 0) and abnormal labrum ( n = 848) groups. The abnormal group consisted of n = 111 with class 1 (degeneration), n = 437 with class 2 (partial or complete tear), and n = 300 with unclassified injury. Patients were randomly divided into training, validation, and test cohort according to the ratio of 55%:15%:30%. Field Strength/Sequence Fat‐saturation proton density‐weighted fast spin‐echo sequence at 3. 0 T . Assessment Convolutional neural network‐6 ( CNN ‐6) was used to extract, discriminate, and detect oblique coronal ( OCOR ) and oblique sagittal ( OSAG ) images. Mask R‐CNN was used for segmentation. LeNet ‐5 was used to diagnose and classify labral injuries. The weighting method combined the models of OCOR and OSAG . The output–input connection was used to correlate the whole diagnosis/classification system. Four radiologists performed subjective diagnoses to obtain the diagnosis results. Statistical Tests CNN ‐6 and LeNet ‐5 were evaluated by area under the receiver operating characteristic ( ROC ) curve and related parameters. The mean average precision ( MAP ) evaluated the Mask R‐CNN . McNemar 's test was used to compare the radiologists and models. A P value < 0.05 was considered statistically significant. Results The area under the curve ( AUC ) of CNN ‐6 was 0.99 for extraction, discrimination, and detection. MAP values of Mask R‐CNN for OCOR and OSAG image segmentation were 0.96 and 0.99. The accuracies of LeNet ‐5 in the diagnosis and classification were 0.94/0.94 ( OCOR ) and 0.92/0.91 ( OSAG ), respectively. The accuracy of the weighted models in the diagnosis and classification were 0.94 and 0.97, respectively. The accuracies of radiologists in the diagnosis and classification of labrum injuries ranged from 0.85 to 0.92 and 0.78 to 0.94, respectively. Data Conclusion Deep learning can assist radiologists in diagnosing and classifying labrum injuries. Evidence Level 3 Technical Efficacy Stage 2