To develop an intelligent model for uterine anatomical structure detection and segmentation (IMSU) by integrating the Efficient Multi-scale Attention (EMA) mechanism into the You Only Look Once version 8 (YOLOv8) framework and evaluate its performance.

 A total of 4326 non-pregnant mid-sagittal uterine ultrasound images were retrospectively collected from Shenzhen Maternity and Child Healthcare Hospital (January 2021-December 2022). Three key anatomical structures (uterine corpus, cervix, and endometrium) were manually annotated to establish an image database. The dataset was divided into training (3460 images), validation (433 images), and test sets (424 images) at an 8∶1∶1 ratio. An IMSU model was constructed by enhancing YOLOv8 with the EMA module. Both the baseline YOLOv8 and IMSU models were trained and validated, followed by performance evaluation on the test set for automated detection and segmentation of uterine structures. Metrics included precision, recall, and mean Average Precision (mAP) at two levels: mAP@50 and mAP@50-95.

 For detection tasks, IMSU outperformed YOLOv8 in overall precision (0.920 vs 0.905), recall (0.939 vs 0.917), and mAP@50 (0.952 vs 0.933). Notably, cervical detection mAP@50 improved from 0.858 to 0.919 and recall increased from 0.778 to 0.842. In segmentation tasks, IMSU achieved higher precision (0.914 vs 0.905), recall (0.933 vs 0.915), mAP@50 (0.952 vs 0.930), and mAP@50-95 (0.677 vs 0.661). Cervical segmentation mAP@50-95 rose from 0.570 to 0.597.

 The EMA-enhanced IMSU significantly improves automated detection and segmentation accuracy for key uterine structures in mid-sagittal ultrasound images, providing technical support for intelligent quantitative uterine measurements and ultrasound-assisted diagnosis with promising clinical applicability.

To evaluate the value of the MobileNetV4 model in identifying aortic arch and its branch anomalies in prenatal ultrasound images.

A total of 12 284 prenatal ultrasound images of vascular rings diagnosed at Shenzhen Maternity and Child Healthcare Hospital from March 2019 to March 2024 were retrospectively collected, including 8913 images with vascular ring anomalies and 3371 normal control images. The images were divided into training, testing, and validation sets at a ratio of 8∶1∶1 according to disease type. A deep learning model based on MobileNetV4 was constructed for the automatic diagnosis of aortic arch branch anomalies, and its performance was compared with five mainstream models applied in different scenarios, including MSPAnet-50, MedMamba, RepVit, ResNet-50, and ViT. Evaluation metrics included accuracy, sensitivity, specificity, precision, area under the receiver operating characteristic (ROC) curve (AUC), and F1 score. The model's comprehensive performance in diagnosing vascular ring anomalies was assessed from multiple perspectives by plotting ROC curves, confusion matrices, Grad-CAM heatmaps, and t-SNE feature visualizations.

In the diagnostic tasks involving double aortic arch (DAA), right aortic arch (RAA), aberrant right subclavian artery (ARSA), and normal images, the MobileNetV4 model demonstrated superior performance, with accuracy, specificity, sensitivity, precision, AUC, and F1 score reaching 94.10%, 98.03%, 94.09%, 94.21%, 99.12%, and 94.04%, respectively. In the heatmaps, only the MobileNetV4 model showed highly concentrated attention on key anatomical structures such as the aortic arch and its branches, exhibiting clearly defined attention regions and sharp boundaries. The confusion matrix revealed that the MobileNetV4 model's predictions were strongly clustered along the diagonal for all categories, indicating high overall diagnostic accuracy, with 100% accuracy achieved on normal control samples. The ROC curves showed that the AUC values for DAA, RAA, ARSA, and normal images were 99.49%, 98.47%, 100%, and 98.51%, respectively, outperforming other comparison models. In the t-SNE visualization, the feature distribution of MobileNetV4 presented more distinct clustering boundaries.

The MobileNetV4 model demonstrates excellent performance in the intelligent recognition of aortic arch branch anomalies in ultrasound images, providing effective technical support for prenatal ultrasound examinations and promoting the intelligent screening and standardized workflow of vascular ring anomalies.

To establish a neonatal birth weight prediction model based on the XGBoost algorithm to provide a precise intelligent tool for clinical perinatal decision-making.

A retrospective analysis was conducted on data from 1018 singleton pregnancies who underwent routine prenatal examinations and delivered at the Yunnan Maternal and Child Health Hospital between January 2021 and April 2023. The dataset included maternal basic information, prenatal ultrasound examination parameters (53 variables in total) during the first, second, and third trimesters, as well as within 7 days before delivery, and neonatal data (gestational age at delivery and birth weight). An XGBoost prediction model was constructed, and its performance was evaluated using multidimensional metrics (mean percentage error, standard deviation of percentage error, mean absolute percentage error, root mean square error, mean absolute error, and clinical prediction accuracy). The XGBoost model was compared with 15 traditional prediction formulas and machine learning models (LightGBM, Logistic, and CatBoost). Feature importance analysis was also performed.

The XGBoost model demonstrated strong performance in predicting neonatal birth weight, with a percentage error standard deviation of 8.21%, mean absolute percentage error of 6.34%, root mean square error of 245.14 g, and mean absolute error of 194.39 g, all error metrics were lower than those of other prediction methods. The clinical accuracy rate reached 71.54%, surpassing those of other predictive approaches. Compared to other benchmark models, the XGBoost model exhibited relatively balanced performance with an overestimation rate of 13.73% and an underestimation rate of 14.71%, both lower than those of Logistic (23.53%/28.43%), LightGBM (25.98%/24.02%), and CatBoost (17.16%/17.65%). Feature importance analysis identified 21 key predictors, among which fetal abdominal circumference before delivery, gestational age at delivery, and maximum amniotic fluid depth before delivery emerged as the most significant predictive variables (feature importance scores of 35, 22, and 8, respectively).

An XGBoost-based temporally correlated neonatal weight prediction model with robust stability and precision has been successfully constructed, and it outperforms traditional regression-based fetal weight prediction methods.

To analyze and summarize the clinical characteristics and ultrasonographic features of decidualized ovarian endometrioma (DOE) during pregnancy to provide a reference for its clinical diagnosis and treatment.

A retrospective analysis was conducted on the medical records of 11 patients clinically diagnosed with DOE during pregnancy at the General Hospital of the PLA from July 2021 to November 2024. The data included clinical information, ultrasound images, follow-up records, and surgical pathological results. The ultrasonographic characteristics and key points for clinical diagnosis and differentiation of DOE during pregnancy were summarized.

All patients exhibited ovarian cysts during pregnancy on ultrasound. The largest diameter of the endometriomas ranged from 3.1 to 7.3 cm, 90.9% (10/11) showed poor echogenicity of intracystic fluid, and the number of papillary projections ranged from 1-6 per cyst. Color Doppler flow imaging detected blood flow signals in 10 cases (flow scores: 2-3 in 72.7% [8/11]), and 72.7% (8/11) had a negative "sliding uterus sign". Three patients underwent surgical resection and had histologically confirmed benign DOE. Sign of decidualization disappeared in 8 cases during the follow-up period. Among these, one case had histologically confirmed benign DOE after surgery due to cyst rupture one year after pregnancy, The remaining 2 cases after abortion and 5 cases during pregnancy showed morphological changes in the cysts in the follow-up period, leading to a diagnosis of DOE based on clinical data.

DOE during pregnancy exhibits characteristic ultrasonographic features. Combining clinical characteristics and follow-up changes can improve diagnostic accuracy and avoid unnecessary surgical intervention. A dynamic follow-up observation strategy is recommended for suspected cases.

To evaluate the value of fetal echocardiography in assessing the severity and prognosis of isolated pulmonary valve stenosis (PS).

A retrospective analysis was conducted on 79 fetuses with PS diagnosed by fetal echocardiography and confirmed by postnatal echocardiography between January 2019 and January 2024 at the First Affiliated Hospital of Hebei North University and Hebei Reproductive and Obstetrics Hospital. According to the severity of PS on postnatal echocardiography, the fetuses were divided into a mild PS group (n=38), a moderate PS group (n=22), and a severe PS group (n=19). Fetal echocardiographic parameters measured included pulmonary valve velocity (PVV), right-to-left ventricular diameter ratio (RV/LV), tricuspid regurgitation velocity (VTR), cardiothoracic ratio (C/T), main pulmonary artery-to-aorta diameter ratio (MPA/AO), tricuspid valve inflow duration-to-cardiac cycle ratio (TVI/CC), and tricuspid-to-mitral valve annular diameter ratio (TV/MV). Receiver operating characteristic (ROC) curve analysis was performed to assess the predictive performance of these parameters for perinatal outcomes.

Fetal echocardiographic findings showed no significant right ventricular remodeling in the mild PS group, mild remodeling with slightly increased thickness and diameter in the moderate PS group, and severe remodeling with reversed ductus arteriosus (DA) flow in the severe PS group. With increasing PS severity, C/T, PVV, and VTR increased significantly, while MPA/AO, RV/LV, TV/MV, and TVI/CC decreased progressively, with overall differences among the three groups being statistically significant (P<0.05). No reversed DA flow was observed in the mild and moderate groups; in the severe group, 100% (19/19) of fetuses had reversed DA flow. The use of prostaglandin E1 increased significantly with PS severity: 0 in mild, 4.55% (1/22) in moderate, and 57.89% (11/19) in severe cases. Percutaneous balloon pulmonary valvuloplasty rates were significantly higher in the moderate (45.45%, 10/22) and severe PS groups (78.95%, 15/19) compared with the mild group (15.79%, 6/38) (P<0.001). Mortality was 0 in both the mild and moderate groups but 10.53% (2/19) in the severe group. In patients with composite adverse outcomes, PVV was significantly higher, and RV/LV and TV/MV were significantly lower than those in patients with non-composite adverse outcomes (P<0.05). The area under the ROC curve values of PVV, TV/MV, and RV/LV for predicting composite adverse outcomes were 0.824, 0.782, and 0.750, respectively; the AUC of the three parameters combined was 0.919.

Fetal echocardiographic parameters can effectively assess the severity of isolated PS, predict perinatal outcomes. facilitate early identification of high-risk fetuses, and provide a scientific basis for individualized perinatal management and intervention strategies, thereby improving fetal prognosis.

To explore the ultrasound imaging features of torsion of the pedicle of old ovarian cysts in infants and young children.

A retrospective study was conducted on 60 infants or young children under the age of 1 year who were diagnosed with ovarian cysts and admitted to the Children's Hospital Affiliated to Zhejiang University School of Medicine from January 2021 to July 2024. The ultrasound features and pathological manifestations were analyzed, and the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of ultrasound features for diagnosing torsion of the pedicle of old ovarian cysts were calculated to evaluate their diagnostic efficacy.

Out of 60 cases of ovarian cysts in infants and young children, 34 cases were old ovarian cyst torsion. The main ultrasound features and their diagnostic efficacy were: calcification (sensitivity: 67.6%; specificity: 80.8%; positive predictive value: 82.1%; negative predictive value: 65.6%); intracystic layering (sensitivity: 67.6%; specificity: 73.1%; positive predictive value: 76.7%; negative predictive value: 63.3%); and vortex sign-like pedicle structure (sensitivity: 47.1%; specificity: 100%; positive predictive value: 100%; negative predictive value: 59.1%). The diagnostic sensitivity, specificity, positive predictive value, and negative predictive value were 26.5%, 100%, 100%, and 51.0%, respectively, when the three characteristics were present simultaneously. The accuracy of ultrasound diagnosis of torsion of old ovarian cyst pedicle was 88.2% (30/34).

The torsion of the pedicle of old ovarian cysts in infants and young children has characteristic ultrasound manifestations, including calcification, intracystic layering, and vortex sign-like pedicle structures, which have good diagnostic value.

To explore the value of artificial intelligence in ultrasound screening of congenital heart disease in children.

A total of 8543 static color Doppler ultrasound images of normal and abnormal cardiac structures from the Children's Hospital Affiliated to Zhejiang University School of Medicine from September 2021 to February 2022 were selected and divided into a training set of 6871 images, a validation set of 833 images, and a test set of 839 images at a ratio of 8:1:1. The OBICnet model was constructed, and its recognition performance was evaluated using F1 score, accuracy, precision, recall, specificity, misdiagnosis rate, and missed diagnosis rate. The performance of the OBICnet model was compared with that of the ResNet50 and GBCnet models. Additionally, 350 static color Doppler images of 50 children with abnormal cardiac structures who visited the Children's Hospital Affiliated to Zhejiang University School of Medicine from November 2022 to January 2023 were collected as external validation data. Twenty-one grassroots ultrasound physicians were divided into three groups based on their years of ultrasound experience: junior, intermediate, and senior. The recognition performance was compared between the OBICnet model and the three groups of physicians on the external validation set.

Compared with ResNet50 and GBCnet models, the OBICnet model had the best recognition performance, with the highest F1 score, accuracy, and precision, which were 97.0%, 98.3%, and 96.7%, respectively, and the lowest misdiagnosis rate and missed diagnosis rate, which were 1.6% and 3.0%, respectively. In the external validation set, the OBICnet model's recognition accuracy, precision, and specificity for normal and abnormal images were 94.6%, 91.4%, and 97.6%, respectively. The recognition accuracy of senior, intermediate, and junior physicians was 89.4%, 83.3%, and 67.8%, respectively, the precision was 78.9%, 73.5%, and 39.3%, respectively, and the specificity was 91.5%, 91.6%, and 81.4%, respectively. The differences in recognition performance between the OBICnet model and the physician groups were statistically significant (all adjusted P<0.017). The recognition sensitivity of the OBICnet model for normal and abnormal images was higher than that of the intermediate and junior physician groups (88.5% vs 61.5% and 31.8%, adjusted P<0.017). The missed diagnosis rate and misdiagnosis rate of the OBICnet model were 11.5% and 2.4%, respectively, which were significantly lower than those of the three groups of physicians (all adjusted P<0.017).

The OBICnet model exhibits superior recognition performance for normal and abnormal cardiac ultrasound images, and it has appreciated application value in the ultrasound screening of congenital heart disease in children.

To explore the differences in clinicopathological characteristics and contrast-enhanced ultrasound (CEUS) manifestations between dual-phenotype hepatocellular carcinoma (DPHCC) and typical hepatocellular carcinoma (THCC), and to identify independent risk factors for DPHCC.

This retrospective study enrolled 71 patients with pathologically confirmed hepatocellular carcinoma after surgical resection (31 cases of DPHCC and 40 cases of THCC) from January 2019 to June 2024. Clinicopathological parameters and CEUS features were compared between the two groups. Multivariate logistic regression was used to identify independent risk factors for DPHCC.

The DPHCC group showed significantly higher rates of serum alpha-fetoprotein (AFP) ≥400 ng/ml (48.4% vs 15.0%, P=0.002) and microvascular invasion (77.4% vs 17.5%, P<0.001) compared to THCC. On CEUS, DPHCC predominantly exhibited heterogeneous enhancement (38.7% vs 15.0%) and rim-like enhancement (16.1% vs 2.5%) in the arterial phase (P=0.003), with earlier contrast agent washout initiation (91.94±37.58 s vs 131.65±59.71 s, P=0.002). Multivariate analysis identified tumor size (odds ratio [OR]=1.019, P=0.002), AFP≥400 ng/ml (OR=2.798, P=0.032), ill-defined margin (OR=3.204, P=0.020), and washout time <120 s (OR=2.221, P=0.049) as independent risk factors for DPHCC.

Multi-parametric CEUS provides critical evidence for noninvasive diagnosis of DPHCC, effectively differentiating it from THCC and optimizing personalized clinical management.

To make a comprehensive analysis of liver stiffness and left ventricular functional changes and their correlations in cirrhosis patients with preserved ejection fraction across Child-Pugh classes using two-dimensional shear wave elastography (2D-SWE) and myocardial work (MW).

A total of 110 cirrhosis patients with preserved left ventricular ejection fraction (LVEF) were recruited from Wenjiang District People's Hospital, Chengdu between August 2023 and December 2024. According to Child-Pugh classification, the patients were stratified into Group A (n=34), Group B (n=37), and Group C (n=39). Thirty-six age- and sex-matched healthy volunteers served as controls. Demographics, blood biochemical parameters, liver stiffness measurements (LSM), conventional echocardiographic indices, 2D-SWE, and MW parameters were collected. Intergroup differences were compared and correlations among parameters were analyzed.

LSM was higher in Group A than in controls (P<0.001). Group B showed higher LSM than both controls and Group A (P<0.001). Group C exhibited higher LSM than controls, Group A, and Group B (P<0.05). In terms of LV diastolic function parameters, e' was decreased in Groups A and B compared to the control group (P<0.001), and demonstrated a further reduction in Group C compared to the control group and groups A and B (P<0.001). E/A ratio was lower in Group B than in controls (P<0.001), and was further reduced in Group C compared to the control group and groups A and B (P<0.001). e'/a' ratio was reduced in Group B compared to controls (P<0.001), with Group C showing a more significant decrease compared to the control group and groups A and B (P<0.001). In terms of left ventricular systolic function parameters, LVEF was significantly higher in Groups B and C compared to the control group (P<0.05). The absolute value of global longitudinal strain (GLS) was significantly increased in Group B compared to both the control group and Group A (P<0.001). The absolute GLS value was significantly lower in Group C than in Group B (P<0.001). Peak strain dispersion (PSD) was significantly increased in Group C compared to the control group and Group A (P<0.001). Global constructive work (GCW) was significantly higher in Group B than in controls (P<0.001). GCW was significantly lower in Group C than in Group B (P<0.001). Global work index (GWI) and global work efficiency (GWE) were significantly reduced in Group C compared to the control group, Group A, and Group B (P<0.001). Global wasted work (GWW) was significantly elevated in Group C compared to the control group, Group A, and Group B (P<0.001). Child-Pugh grade correlated positively with LSM, AST, TBA, and GWW (r= 0.872, 0.499, 0.533, and 0.446, respectively; P<0.001), and negatively with DBP, E/A, e', e'/a', and GWE (r=−0.483, −0.562, −0.669, −0.659, and −0.479, respectively; P<0.001). LSM showed positive correlations with CO, AST, and TBA (r=0.467, 0.584, and 0.585, respectively; P<0.001), and negative correlations with DBP, E/A, e', and e'/a' (r=−0.513, −0.491, −0.542, and −0.571, respectively; P<0.001).

With the worsening of Child-Pugh classification in liver cirrhosis, liver dysfunction deteriorates, accompanied by increased LSM values and heightened measurement dispersion, left ventricular diastolic function declines, and systolic function initially exhibits compensatory enhancement but subsequently undergoes decompensatory reduction. Liver-heart functional impairment demonstrates synergistic effects. The combined use of 2D-SWE and MW techniques provided a more comprehensive and effective assessment of liver disease severity and left ventricular functional changes in cirrhotic patients, which offered crucial imaging evidence for targeted clinical interventions and improving patient prognosis.