ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

Purpose: The aim of this study was to investigate the feasibility of an intelligent handheld ultrasound (US) device for assisting non-expert general practitioners (GPs) in detecting carotid plaques (CPs) in community populations. Methods: This prospective parallel controlled trial recruited 111 consecutive community residents. All of them underwent examinations by non-expert GPs and specialist doctors using handheld US devices (setting A, setting B, and setting C). The results of setting C with specialist doctors were considered the gold standard. Carotid intima-media thickness (CIMT) and the features of CPs were measured and recorded. The diagnostic performance of GPs in distinguishing CPs was evaluated using a receiver operating characteristic curve. Inter-observer agreement was compared using the intragroup correlation coefficient (ICC). Questionnaires were completed to evaluate clinical benefits. Results: Among the 111 community residents, 80, 96, and 112 CPs were detected in settings A, B, and C, respectively. Setting B exhibited better diagnostic performance than setting A for detecting CPs (area under the curve, 0.856 vs. 0.749; P<0.01). Setting B had better consistency with setting C than setting A in CIMT measurement and the assessment of CPs (ICC, 0.731 to 0.923). Moreover, measurements in setting B required less time than the other two settings (44.59 seconds vs. 108.87 seconds vs. 126.13 seconds, both P<0.01). Conclusion Using an intelligent handheld US device, GPs can perform CP screening and achieve a diagnostic capability comparable to that of specialist doctors.

Objective To analyze the application of in-use laparoscopes in clincal departments to enhance the laparoscope's effectiveness.Methods The effective utilization data of 29 laparoscopes from January 2024 to June 2024 were acquired with an hospital intelligent medical device management platform.Comparisons were carried out in terms of average daily workload,average daily hours of use and average daily efficiency between the laparoscopes from different departments and brands with non-parametric Kruskal-Wallis test and between the conventional and 3D laparoscopes with non-parametric Mann-Whitney U test.Pearson's correlation coefficient was used to analyze the correlation between the number of years of laparoscope use and the average daily workload,the average daily hours of use,and the average daily efficiency and the chi-square test was applied to investigating the relationship between the surgery grade and the type of equipment selected.Results Thoracic sugery department had the highest average daily workload(3.82 person-time),while the general medical department had the highest average daily hours of use(443.76 min)and the highest daily efficiency(92.45％).There were significant differences between the laparoscopes from different departments in average daily workload,average daily hours of use and average daily efficiency(P＜0.05).Brand D laparoscope behaved the best among brands of laparoscopes with the highest average daily workload(3.72 person-time),average daily hours of use(394.41 min)and average daily efficiency(82.17％).There were sig-nificant differences between the brands of laparoscopesin average daily workload,average daily hours of use and average daily efficiency(P＜0.05).3D laparoscopes obviously gained advantages over the conventional ones in average daily workload,average daily hours of use and average daily efficiency(P＜0.05).The number of years of use correlated negatively with average daily workload,average daily hours of use and average daily efficiency,with Pearson correlation coefficients being-0.095,-0.039 and-0.039 respectively.Grade Three and Four surgeries had significant differences in types of selected equipment(P＜0.001),and 3D laparoscopes were preferred for Grade Four surgery.Conclusion Utilization analysis of laparoscopes provides data support for optimized application and setup of laparoscopes.[Chinese Medical Equipment Journal,2025,46(2):87-91]

Objective To explore the relationship between community health follow-up management and the mental health of the long-term care insurance residents,and to provide a basis for the construction of an integrated community home care service mode for disabled elders.Methods The residents were selected through cluster sampling who participated in LTCI home care from Jan 1 to Dec 31,2021.After a year of participation,the subjects'mental health was assessed face-to-face by trained community doctors using the Self-Rating Anxiety Scale and the Self-Rating Depression Scale.By referring to residents'electronic health records combined with on-site questionnaire survey,community doctors collected the demographic information and health follow-up management provided by primary medical and health institutions.The multivariate logistic regression were conducted to evaluate the association between follow-up care and mental health outcomes.Results The study consisted of 399 LTCI-enrolled individuals,57.64%(n=230)received follow-up care by family physicians.The prevalence of anxiety and depression among participants was 19.80%(n=79)and 67.67%(n=270),respectively.Univariate analysis found that community health follow-up management could underscore the potential impact of follow-up care in mitigating anxiety(χ2=38.926,P<0.001)and depression(χ2=14.598,P<0.001)among LTCI enrollees.Multivariate analysis revealed that follow-up care was an independent protective factor against anxiety(adjusted OR=0.351,95%CI:0.176-0.701,P=0.003).However,follow-up care did not significantly impact depression prevalence.Additionally,LTCI grade and education level were also identified factors influencing the mental health of participants(P<0.05).Conclusion Community health service centers provide health follow-up management that plays a positive role in alleviating the anxiety symptoms of disabled residents under long-term care insurance home care.It is an effective way to improve the quality of LTCI home care services.

Aim To observe the effect of lipopolysac-charide(LPS)induced supernatant of BV-2 cells on the inflammatory response and apoptosis of HT22 neu-rons.Methods After the concentration and time of LPS were determined by CCK-8 method,BV-2 cells were cultured with medium without LPS and medium containing LPS,the morphological changes of BV-2 microglia were observed by inverted microscope,and the CD86/CD206 ratio of BV-2 microglia was detected by immunofluorescence.Subsequently,BV-2 cell cul-ture supernatants were isolated and added to HT22 neuronal culture to observe the effect on the inflamma-tory response of HT22 neurons.The proliferation of HT22 neurons was detected by CCK-8 method and EdU method.The structural changes of HT22 neurons were observed under the microscope and examined by urani-um-lead staining.The levels of cytokines interleukin-1β(IL-1β),interleukin-10(IL-10),nuclear factor kappa-B(NF-κB)and tumor necrosis factor-α(TNF-α)were detected by enzyme-linked immunosorbent as-say(Elisa).Neuronal apoptosis was detected by the TUNEL method.The protein expressions of Bax,Bcl-2 and inflammatory factors were detected by Western blot.Results After induction with 1 mg·L-1 LPS,BV-2 cells exhibited increased cell body size,thicker protrusions on both side,and some cells showed de-formed protrusions,the CD86/CD206 ratio in BV-2 cells decreased,promoting the transformation of BV-2 cells from M2 type to M1 type.After treating with the culture supernatant of BV-2 cells,HT22 neuronal cell activity and proliferation were reduced,axons short-ened,and the number of cells decreased.Neuronal cell bodies were enlarged and some cells were de-formed,with damaged cell membranes,round cell nu-clei but displaced nucleoli from the normal position,swollen mitochondria with vacuoles,reduced internal ridge structures,and increased levels of inflammatory factors NF-κB,IL-1 β,and TNF-α(P＜0.05 or P＜0.01),while the anti-inflammatory factor IL-10 de-creased(P＜0.05),protein expression of the pro-apoptotic indicator Bax increased(P＜0.01),and the protein expression of the anti-apoptotic indicator Bcl-2 decreased(P＜0.05).Conclusion After induction of BV-2 cell polarization by LPS,the supernatant could inhibit HT22 neuronal cell viability,upregulate inflam-matory factor expression and promote apoptosis.

Objective To develop a novel type of transparent simulation manikin as a surgical training model to meet the surgical treatment demand on the medical train.Methods A transparent manikin was developed with the steps of basic data collection,motherboard design and manufacture and module production and assembly.Firstly,basic data collection was carried out with reference to standardized human anatomy and parameters.Secondly,some software such as UG NX7.5 was used to construct the motherboard of the manikin.Finally,module production and assembly were performed with the materials of acrylic,transparent rubber,silicone and hydrogel and the technology of silicone infusion.Results The transparent manikin developed had its anatomy structure close to that of the real body and high visuality for its internal and external components,which simulated a variety of war wounds and thus could be integrated with the surgical training scenarios on the medical train effectively.Conclusion The transparent manikin developed is characterized by high visuality,modularity and blood flow,and meets the demands for surgical training on the medical train.[Chinese Medical Equipment Journal,2025,46(6):111-115]

Objective To investigate the stability and reproducibility of the treatment fractions during the intensity-modulated radiotherapy(IMRT)for breast cancer and the effect of respiratory motion on the dose irradiation of breast cancer radiotherapy by comparing the results of breast cancer dosimetric verification based on fractionated images by an electronic portal imaging device(EPID).Methods A total of 28 IMRT patients admitted to some hospital from January to June 2023 were grouped according to the pathological results and effects of respiratory motion on the accuracy of radiotherapy during clinical treatment,including 14 cases in a breast group and 14 cases in a non-breast group with 8 ones of head and neck tumors,5 ones of esophageal cancer and 1 case of cervical cancer.All the patients underwent a scan with cone beam computed tomography(CBCT)before the first radiotherapy,and image registration was carried out with a positioning CT.An EPID was used to acquire transmission dose images of 10 fractions of radiotherapy,and γ analysis was performed using the RIT 113 QA software to compare the images of the subsequent 9 fractions with those of the first fraction,with the images of the first fraction of radiotherapy as the baseline values.Absolute maximum dose normalization was implemented under the condition of 10％dose assessment threshold,and the γ-pass rates under the 3 criteria of 2％/2 mm,3％/2 mm and 3％/3 mm were counted separately.The fraction dose verification results of the 28 patients were divided into 3 treatment phases of 2-4 times(T1),5-7 times(T2)and 8-10 times(T3)to analyze the stability of dose irradiation during the radiotherapy.SPSS 22.0 software was used for statistical analysis.Results Under the condition of 10％dose assessment threshold,the breast and non-breast groups had the γ-pass rates being(95.80±2.65)％and(94.60±6.59)％under the 2％/2 mm criterion and(98.46±1.31)％and(97.50±3.30)％under the 3％/2 mm criterion respectively,and the differences were statistically significant(all P<0.05).Under the assessment criteria of 2％/2 mm,3％/2 mm and 3％/3 mm,the breast group had the γ-pass rates of fractions of treatment significantly lower than those of the non-breast group(all P<0.05),while the γ-pass rates showed no significant differences at T1,T2 and T3 treatment phases(all P>0.05).Conclusion EPID fraction images contribute to evaluating IMRT accuracy effectively.IMRT has high stability and reproducibility during the treatment cycle,while respiration may result in dose deviation during the fraction radiotherapy for breast cancer,and optical surface tracking technology or active breathing control technology is suggested to be involved in to relieve dose deviation.[Chinese Medical Equipment Journal,2025,46(6):54-58]

Objective To explore a deep learning-based automatic bone age estimation model for elbow joint X-ray images of Chinese Han adolescents and children and evaluate its performance.Methods A total of 943(517 males and 426 females)elbow joint frontal view X-ray images of Chinese Han ado-lescents and children aged 6.00 to<16.00 years were collected from East,South,Central and North-west China.Three experimental schemes were adopted for bone age estimation.Scheme 1:Directly in-put preprocessed images into the regression model;Scheme 2:Train a segmentation network using"key elbow joint bone annotations"as labels,then input segmented images into the regression model;Scheme 3:Train a segmentation network using"full elbow joint bone annotations"as labels,then in-put segmented images into the regression model.For segmentation,the optimal model was selected from U-Net,UNet++and TransUNet.For regression,VGG16,VGG19,InceptionV2,InceptionV3,ResNet34,ResNet50,ResNet101 and DenseNet121 models were selected for bone age estimation.The dataset was randomly split into 80%(754 samples)for training and validation for model fitting and hyperparameter tuning,and 20%(189 samples)as an internal test set to test the performance of the trained model.An additional 104 elbow joint X-ray images from the same demographic and age group were col-lected and used as an external test set.Model performance was evaluated by comparing the mean ab-solute error(MAE),root mean square error(RMSE),accuracies within±0.7 years(P±0.7 years)and±1.0 years(P±1.0 years)between the estimated age and the actual age,and by drawing radar charts,scat-ter plots,and heatmaps.Results When segmented with Scheme 3,the UNet++model achieved good segmentation performance with a segmentation loss of 0.000 4 and an accuracy of 93.8%at a learning rate of 0.000 1.In the internal test set,the DenseNet121 model with Scheme 3 yielded the best results with MAE,P±0.7 years and P±1.0 years being 0.83 years,70.03%,and 84.30%,respectively.In the external test set,the DenseNet121 model with Scheme 3 also performed best,with an average MAE of 0.89 years and an average RMSE of 1.00 years.Conclusion When performing automatic bone age estima-tion using elbow joint X-ray images in Chinese Han adolescents and children,it is recommended to use the UNet++model for segmentation.The DenseNet121 model with Scheme 3 achieves optimal per-formance.Using segmentation networks,especially that trained with annotation areas encompassing the full elbow joint including the distal humerus,proximal radius,and proximal ulna,can improve the ac-curacy of bone age estimation based on elbow joint X-ray images.

Objective To explore a deep learning network model suitable for bone age estimation using shoulder joint X-ray images in Chinese Han adolescents.Methods A retrospective collection of 1 286 shoulder joint X-ray images of Chinese Han adolescents aged 12.0 to<18.0 years(708 males and 578 females)was conducted.Using random sampling,approximately 80%of the samples(1 032 cases)were selected as the training and validation sets for model learning,selection and optimization,and the other 20%samples(254 cases)were used as the test set to evaluate the model's generalization ability.The original single-channel shoulder joint X-ray images and dual-channel inputs combining original images with segmentation labels(manually annotated shoulder joint regions multiplied pixel-by-pixel with original images,followed by segmentation via the U-Net++network to retain only key shoulder joint region information)were respectively input into four network models,namely VGG16,ResNet18,ResNet50 and DenseNet121 for bone age estimation.Additionally,manual bone age estimation was con-ducted on the test set data,and the results were compared with the four network models.The mean absolute error(MAE),root mean square error(RMSE),coefficient of determination(R2),and Pear-son correlation coefficient(PCC)were used as main evaluation indicators.Results In the test set,the bone age estimation results of the four models with dual-channel input of shoulder joint X-ray images outperformed those with single-channel input in all four evaluation indicators.Among them,DenseNet121 with dual-channel input achieved best results with MAE of 0.54 years,RMSE of 0.82 years,R2 of 0.76,and PCC(r)of 0.88.Manual estimation yielded an MAE of 0.82 years,ranking second only to dual-channel DenseNet121.Conclusion The DenseNet121 model with dual-channel input combined with original images and segmentation labels is superior to manual evaluation results,and can effectively estimate the bone age of Chinese Han adolescents.