To develop a ConvNeXt deep learning model based on hepatic ultrasound radiofrequency (RF) signals and the ultrasound attenuation parameter (UAP), using liver biopsy as the reference standard, and to evaluate its diagnostic performance for significant hepatic steatosis in liver diseases.

A total of 1222 patients with liver diseases who underwent evaluation at Beijing Ditan Hospital, Capital Medical University between January 2020 and February 2025 were retrospectively analyzed. RF signals, UAP measurements, and liver biopsy data were collected. The dataset was divided into a training set (n=851), a validation set (n=192), and a test set (n=179). RF signal acquisition and UAP measurement were performed prior to liver biopsy, and corresponding histopathological findings were obtained. Hepatic steatosis graded as F0–F1 was defined as negative (label 0), while ≥F2 was defined as positive (label 1). To construct an artificial intelligence model for assessing significant hepatic steatosis, each column of ultrasound RF signal data was normalized using Z-score standardization. ConvNeXt-Tiny was used as the backbone network, initialized with pretrained weights from the ImageNet dataset, and the mean of the first-layer three-channel weights was used to initialize single-channel weights. UAP was incorporated as a scalar feature and concatenated with image features extracted by the ConvNeXt backbone at the final linear layer. The model prediction score was generated through a fully connected layer followed by a Sigmoid function, and final classification was determined using a threshold. Receiver operating characteristic (ROC) curves were generated to evaluate the performance of the ConvNeXt model and UAP for detecting significant hepatic steatosis, and their performance was compared.

In the test set, the area under the ROC curve (AUC) of the ConvNeXt model for diagnosing significant hepatic steatosis was 0.838 (95% confidence interval [CI]: 0.769–0.906). Using the cutoff value (0.225) determined in the validation set, the confusion matrix [[True Negative (TN) False Positive (FP) ][False Negative (FN)True Positive (TP)]] in the test set was [[108 36][6 29]], yielding a sensitivity of 82.86% and specificity of 75.00%. In addition, based on the ROC curve in the test set, the optimal cutoff value was 0.272, at which the confusion matrix [[TN FP ][FN TP]] was [[115 29][6 29]], yielding a sensitivity of 82.86% and specificity of 79.86%. For UAP, the AUC for diagnosing significant hepatic steatosis was 0.802 (95%CI: 0.723–0.881), with a threshold of 269 dB/m, and the sensitivity and specificity were 74.29% and 72.22%, respectively.

The ConvNeXt deep learning model based on RF signals and UAP demonstrates favorable performance in identifying significant hepatic steatosis (≥F2) in liver diseases, and may serve as a useful tool for clinical screening and follow-up of patients with steatotic liver disease.

To evaluate the value of ultrasound derived fat fraction (UDFF) in the quantitative assessment of hepatic steatosis in patients with varying degrees of hypertriglyceridemia (HTG).

A total of 156 HTG patients and 40 healthy controls who visited Xiyuan Hospital, China Academy of Chinese Medical Sciences from October 2023 to December 2024 were enrolled. Clinical data, total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), ultrasound grade of fatty liver, subcutaneous fat thickness (SFT), visceral fat thickness (VFT), and UDFF were collected. HTG patients were divided into two groups based on TG levels: borderline elevated TG group (1.7≤TG＜2.3 mmol/L, n=63) and elevated TG group (2.3≤TG＜5.7 mmol/L, n=93). The intra-observer consistency of UDFF measurements was evaluated, differences in UDFF were compared among groups, and its correlation with clinical indicators was analyzed.

The intra-observer consistency of UDFF measurements was excellent (intra-class correlation coefficient=0.971, P＜0.001). The UDFF value in the HTG group was significantly higher than that of the control group (11.0% vs 5.0%, P＜0.001). UDFF showed an increasing trend with rising TG levels: control group (5.0%) ＜borderline elevated TG group (6.0%)＜elevated TG group (15.0%), with statistically significant differences in pairwise comparisons (P＜0.05). In HTG patients, UDFF was positively correlated with fatty liver grade, TG, and VFT (r=0.817, 0.745, and 0.641, respectively; P＜0.001) and negatively correlated with HDL-C (r=-0.546, P＜0.001). Multiple linear regression analysis revealed that fatty liver grade (β=0.390), TG (β=0.402), VFT (β=0.170), and SFT (β=0.121) were independent risk factors for UDFF (R²=0.819).

UDFF is significantly correlated with the grade of fatty liver and metabolic indicators in HTG patients, enabling early identification and quantitative assessment of hepatic steatosis in this population.

To develop a preoperative predictive model for central compartment lymph node metastasis (CLNM) in papillary thyroid carcinoma (PTC) by integrating ultrasound radiomics features, deep learning features, and clinical characteristics and to evaluate its performance.

A retrospective analysis was conducted on 510 pathologically confirmed PTC patients who underwent surgical treatment at Fudan University Shanghai Cancer Center between December 2020 and March 2023. All patients received ultrasound-guided fine-needle aspiration biopsy preoperatively. Postoperative pathology identified 387 CLNM-negative and 123 CLNM-positive cases. Regions of interest were manually delineated on grayscale ultrasound images by experienced radiologists. Radiomics features were extracted using Pyradiomics, and deep learning features were obtained via a pretrained VGG16-BN model. After feature selection and dimensionality reduction, the final radiomics and deep learning features were combined with clinical features to construct a multimodal feature set. All cases were split into training and independent test sets at a 4∶1 ratio. Five-fold cross-validation was applied during training for model construction and hyperparameter optimization, and a support vector machine (SVM)-based predictive model was established. Model performance was evaluated on the independent test set using the area under the receiver operating characteristic curve (AUC) and its 95% confidence interval (CI).

ROC curve analysis showed that the discriminative ability of the radiomics-only model was the weakest (AUC=0.63), while the performance of the deep learning-only model was significantly improved (AUC=0.91). The bimodal fusion model combining deep learning and radiomics further enhanced the performance (AUC=0.92), and the full-feature fusion model (radiomics + deep learning + clinical features) achieved the best cross-validation performance (AUC=0.93). On the independent test set, the trimodal fusion model achieved an AUC of 0.962 (95%CI: 0.915-0.994), a specificity of 0.987 (95%CI: 0.931-0.999), and a sensitivity of 0.583 (95%CI: 0.366-0.779), demonstrating strong overall discriminative ability and high negative exclusion capability.

The multimodal predictive model integrating ultrasound radiomics, deep learning features, and clinical risk factors effectively enhances preoperative risk stratification for CLNM in PTC patients, offering potential clinical decision-support value. Its generalizability warrants further validation in multicenter, large-sample studies.

To evaluate the clinical utility of echocardiography in the prenatal diagnosis of fetal persistent left superior vena cava (PLSVC) with absent right superior vena cava (RSVC).

A retrospective analysis was conducted on the echocardiographic features of 43 fetal cases diagnosed with PLSVC and absent RSVC at Chongqing Health Center for Women and Children from November 2016 and December 2024. Associated genetic testing results were also reviewed and analyzed.

Among the 43 cases included, 24 (24/43, 55.81%) were identified as having isolated PLSVC with absent RSVC, 8 (8/43, 18.60%) were associated with additional intracardiac structural anomalies, and 11 (11/43, 25.58%) exhibited both intracardiac and extracardiac malformations. Genetic testing was performed in 32 cases (32/43, 74.42%), of which 30 (30/32, 93.75%) had normal findings. Pregnancy outcomes included 27 live births (27/43, 62.79%) and 12 terminations (12/43, 27.91%). Among the live-born cases, 24 were diagnosed with isolated PLSVC with absent RSVC. Echoardiographic features of isolated PLSVC with absent RSVC included absence of the RSVC, visualization of the left superior vena cava draining into the right atrium via the coronary sinus, and reversed flow from the innominate vein into the left superior vena cava. Three cases were associated with additional structural anomalies. Prenatal ultrasound findings in these three cases included muscular ventricular septal defect, a slightly enlarged left heart compared to the right, and partial atrioventricular septal defect.

Fetal echocardiography is an important method for detecting PLSVC with an absent RSVC and holds significant clinical application value. Fetuses presenting with isolated PLSVC and absent RSVC generally have favorable postnatal outcomes.

To investigate the correlation of fetal aortic pulse wave velocity (PWV) with gestational age and cardiac annulus diameters in the second and third trimesters of pregnancy, and to explore the specific characteristics of fetal intrauterine growth and development.

This prospective study included normal fetuses (gestational age: 18 weeks to 39⁺⁶ weeks) undergoing fetal ultrasound at the Second Xiangya Hospital from October 2024 to March 2025. Dual-gate Doppler ultrasound successfully acquired valid measurements for ascending aortic PWV in 149 cases and descending aortic PWV in 201 cases. PWV was compared between the ascending and descending aorta, and the correlation of ascending and descending aortic PWV with gestational age and diameters of cardiac annuli, including the mitral valve annulus (MV), tricuspid valve annulus (TV), aortic valve annulus (AV), and pulmonary valve annulus (PAV), were analyzed. Multiple regression models were constructed to determine the optimal curve fitting.

Both ascending and descending aortic PWV showed a significant positive correlation with gestational age (r=0.30 and 0.38, respectively, both P＜0.01), with a Sigmoid curve showing the best goodness of fit. Descending aortic PWV was significantly higher than ascending aortic PWV in both the second and third trimesters [second trimester: (2.22±0.44) m/s vs (1.03±0.42) m/s, P＜0.01; third trimester: 2.48 (2.32, 2.66) m/s vs 1.31 (1.10, 1.47) m/s, P＜0.01] Additionally, there was a positive correlation between ascending and descending aortic PWV and each cardiac annulus diameter in normal second and third trimester fetuses (MV: r=0.26 and 0.34; TV: r=0.24 and 0.36; AV: r=0.28 and 0.33; PAV: r=0.28 and 0.39; all P＜0.01).

This study adopted the dual-pulse Doppler technique and demonstrated that fetal aortic PWV increases with gestational age, while also revealing segmental distribution differences in fetal aortic PWV. These findings facilitate further investigation of fetal aortic function and provide a novel reference indicator for the clinical quantitative assessment of fetal aortic elasticity.

To investigate the prenatal ultrasound and clinical features of fetal anomalous connection of the ductus venosus (DV) nd evaluate their clinical significance and value.

A retrospective analysis was conducted on clinical data from 64 fetuses diagnosed with anomalous DV connection by prenatal ultrasound at Chongqing Health Center for Women and Children between January 2017 and June 2024. Prenatal ultrasound features, associated structural anomalies, genetic examinations, and pregnancy outcomes were analyzed.

Among the 64 cases of anomalous connection of the DV included, 56.25% (36/64) had abnormal connection to the inferior vena cava, 9.38% (6/64) demonstrated abnormal connection to the coronary sinus, 21.88% (14/64) exhibited abnormal connection to the hepatic vein, and 12.50% (8/64) showed abnormal connection to the right atrium; 23 (23/64, 35.94%) had isolated anomalous DV connection, 32 had other complex cardiovascular abnormalities, 20 had other extracardiac abnormalities, and 11 had both intracardiac and extracardiac abnormalities; 10 underwent serological screening for Down's syndrome (2 of which were high-risk), and 26 underwent amniocentesis, with 7 cases showing abnormal results; 26 pregnancies were terminated, and 38 resulted in live births. Thirty-seven of the live-born infants showed spontaneous closure of the DV during follow-up with no obvious developmental abnormalities (20 of them had isolated anomalous DV connection). One infant died of pulmonary hemorrhage after birth, with anomalous DV connection combined with large ventricular septal defect and single umbilical artery. Notably, the live birth rate in the isolated DV anomalous connection group (20/23, 86.96%) was significantly higher than that in the group with associated structural anomalies (18/41, 43.90%) (P=0.001).

Prenatal ultrasound can identify anomalous DV connection and evaluate associated structural anomalies, and is of great value in the diagnosis of abnormal DV course. Fetal anomalous DV connection may be complicated by cardiac and extracardiac anomalies and is associated with potential chromosomal abnormalities. Isolated anomalous DV connection has a favorable prognosis.

To evaluate the clinical efficacy of ultrasound-guided intra-articular injection of corticosteroids via the posterior shoulder approach or anterior rotator interval approach for persistent shoulder pain after rotator cuff repair.

A total of 75 patients with shoulder pain following rotator cuff repair admitted to the Department of Orthopedics, the First Medical Center of Chinese People's Liberation Army General Hospital from June 2023 to June 2024 were retrospectively enrolled. According to whether tenotomy of the long head of the biceps brachii was performed intraoperatively, the patients were divided into two groups: Group A (38 cases) underwent rotator cuff repair combined with tenotomy of the long head of the biceps brachii) and received posterior intra-articular shoulder injection, while Group B (37 cases) underwent rotator cuff repair alone and received intra-articular shoulder injection via the anterior rotator interval. The injection regimen consisted of 1 ml diprospan plus 9 ml 0.2% ropivacaine. The visual analogue scale (VAS), Constant–Murley score (CMS), University of California, Los Angeles shoulder score (UCLA), Sports Injury Rehabilitation Adherence Scale (SIRAS), and Likert satisfaction scale were adopted to evaluate clinical outcomes at 1 week, 3 months, and 6 months after injection. Imaging examinations were performed to assess the retear rate of the rotator cuff.

Overall differences in VAS, CMS, and UCLA scores between baseline and postoperative time points were statistically significant (P＜0.05). The preoperative VAS score was 7 (6, 8), which decreased to 4(3, 5) at 1 week postoperatively, 4 (3, 4) at 3 months, and 3 (3, 4) at 6 months. The preoperative CMS and UCLA scores were 49 (45, 55) and 16 (15, 17), respectively, and increased to 80 (80, 82) and 26 (26, 26) at 6 months postoperatively. Compared with baseline, the differences of all scale scores at each follow-up time point after treatment were statistically significant (adjusted P＜0.05). The preoperative SIRAS score was 4 (3, 5), while postoperative SIRAS scores were all ≥13.0, indicating good rehabilitation adherence. The preoperative Likert satisfaction scale score was 3 (3, 3), which improved to 5 (4, 5) at 6 months after operation. There were no significant differences in all scale scores between Group A and Group B before treatment as well as at 1 week, 3 months, and 6 months after treatment (all P＞0.05). Rotator cuff retear occurred in 2 cases in Group A (2/38, 5.2%) and in 1 case in Group B (1/37, 2.7%).

Ultrasound-guided intraarticular corticosteroid injection via both the posterior shoulder and anterior rotator interval approaches are safe and effective in relieving shoulder pain and improving shoulder function after rotator cuff repair.

To construct a novel tumor microenvironment (TME)-responsive nanoplatform, NG@Fe₃O₄-PFH@CM, and to explore its potential to synergize with high-intensity focused ultrasound (HIFU) ablation therapy for triple-negative breast cancer (TNBC) by inducing ferroptosis and activating anti-tumor immunity.

We fabricated a responsive nanoplatform (NG@Fe₃O₄-PFH@CM) using a glutathione/pH dual-responsive nanogel as the framework to encapsulate ferric ion-loaded Fe₃O₄ and the phase-change agent perfluorohexane (PFH), which was then coated with cancer cell membranes (CM). The physicochemical properties of the nanoplatform and its HIFU-triggered acoustic phase-change behavior were systematically characterized. The efficacy of the nanoplatform was evaluated in vitro and in 4T1 breast cancer orthotopic and metastasis models. Assessments included its ability to enhance HIFU-sensitized ultrasound imaging, biocompatibility, tumor accumulation, ferroptosis induction, immune activation, and its synergistic effect with HIFU on inhibiting tumor growth and metastasis.

In vitro experiments confirmed that the nanoplatform exhibited excellent biocompatibility, tumor accumulation capacity, and tumor microenvironment (TME)-responsive drug release profiles. The Fe₃O₄ released within the TME resulted in the generation of substantial reactive oxygen species (ROS) via the Fenton reaction, triggering severe lipid peroxidation and inducing ferroptosis. HIFU combined with NG@Fe₃O₄-PFH@CM effectively induced cellular ferroptosis and significantly promoted dendritic cell (DC) maturation (29.1±0.9%), nearly doubling the rate observed in the Saline group (t=21.13, P＜0.0001). In vivo studies indicated that the HIFU-triggered phase transition significantly enhanced contrast-enhanced ultrasound (CEUS) imaging at the tumor site. Driven by the marked elevation in intratumoral DC maturation (38.5±0.8%, 2.2 times that of the Saline group; t=36.59, P＜0.0001) and the consequent immune activation, the combination therapy achieved a tumor inhibition rate of 86.1% (F=45.63, P＜0.0001) and effectively suppressed lung metastasis (inhibition rate of 90.0%, F=92.96, P＜0.0001).

We have successfully constructed a functional nanoplatform with TME-responsive drug release. This platform not only enhances ultrasound imaging for treatment monitoring through phase change but also effectively boosts tumor ferroptosis and activates cellular immunity, thereby significantly inhibiting tumor proliferation and distant metastasis. This work provides a new strategy for enhancing HIFU efficacy.