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.

Objective To analyze the efficacy, safety, and learning curve of the three-lobe holmium laser enucleation of the prostate(HoLEP)in the treatment of benign prostatic hyperplasia(BPH)in a county-level hospital.Methods A retrospective analysis was conducted on the clinical data of 65 BPH patients who underwent the three-lobe HoLEP performed by a single surgeon at the Department of Urology, Yining County People's Hospital, during Dec.2023 and Jun.2024.The enucleation efficiency was calculated by dividing the weight of the enucleated prostatic tissue by the enucleation time.A case scatter diagram of enucleation efficiency was plotted according to the chronological order of the operations, and the learning curve was analyzed after fitting.Based on the inflection points of the learning curve, the learning process was divided into the initial learning, mastery, and proficiency phases.The basic clinical data, perioperative indicators, postoperative complications, and follow-up indicators were compared among the different learning phases.Results All 65 procedures were successfully completed.Marked by the enucleation efficiency reaching a plateau, cases 1-20 were defined as the initial learning phase, cases 21-40 as the mastery phase, and cases 41 onwards as the proficiency phase.The prostate volume in the proficiency phase was significantly larger than that in the initial learning and mastery phases(both P＜0.05).There were no statistically significant differences in other baseline characteristics among the three groups(all P＞0.05).The operation time [(105.50±19.12)min vs.(85.25±26.92)min vs.(69.00±23.58)min] and enucleation time [(76.90±14.19)min vs.(63.70±22.24)min vs.(48.80±20.48)min] showed a decreasing trend across the three groups(all P＜0.05).The enucleation efficiency in both the mastery and proficiency phases was significantly higher than that in the initial learning phase [(1.16±0.44)g/min vs.(0.85 ±0.25)g/min, P＜0.05;(1.36±0.49)g/min vs.(0.85±0.25)g/min, P＜0.05].The enucleation efficiency in the proficiency phase was slightly higher than that in the mastery phase, but the difference was not statistically significant(P= 0.389).There was no significant difference in the incidence of perioperative complications among the three groups(all P＞0.05).At the 6-month follow-up, the international prostate symptom score(IPSS), post-void residual(PVR), and maximum urinary flow(Qmax)were significantly improved compared to preoperative values in all three groups(all P＜0.05);however, no significant differences were observed among the three groups(all P＞0.05).Conclusion The three-lobe HoLEP for the treatment of BPH is safe and effective in a county-level hospital setting.Surgeons with some experience in endoscopic surgery can preliminarily master this technique after a learning period of approximately 20 procedures.

Objective: To analyze the relationship between parental psychological control and Internet Gaming Disorder (IGD) among junior high school students, so as to provide evidence for preventing IGD development in adolescents. Methods: From August 2019 to February 2020, a survey was conducted among 1 169 junior high school students from three middle schools in Xian using stratified cluster sampling. The Parental Psychological Control Scale and IGD Scale were administered to assess parental psychological control and IGD prevalence. Univariate and binary Logistic regression analyses were used to explore IGD risk factors and their correlation with parental psychological control. Results: The detection rate of IGD in middle school students was 19.9%(184/1 169). Multivariate Logistic regression revealed that compared to those with lower parental psychological control scores(≤21 points), students with higher parental psychological control scores (>21 points) had a higher risk of IGD (OR=1.82, 95%CI=1.21-2.74), a 1.58fold higher risk of selfperceived gaming addiction (95%CI=1.07-2.30), as well as reduced likelihood of seeking external help to reduce gaming time (OR=0.66, 95%CI=0.47-0.94) (P<0.05). Conclusions Parental psychological control may elevate the risks of IGD and selfperceived addiction while diminishing proactive helpseeking behaviors to reduce gaming time. Parents should enhance communication with adolescents and provide positive guidance to mitigate potential gamingrelated harms.

Objective To reveal the scientific output and trends in pulmonary nodules/early-stage lung cancer prediction models. Methods Publications on predictive models of pulmonary nodules/early lung cancer between January 1, 2002 and June 3, 2023 were retrieved and extracted from CNKI, Wanfang, VIP and Web of Science database. CiteSpace 6.1.R3 and VOSviewer 1.6.18 were used to analyze the hotspots and theme trends. Results A marked increase in the number of publications related to pulmonary nodules/early-stage lung cancer prediction models was observed. A total of 12581 authors from 2711 institutions in 64 countries/regions published 2139 documents in 566 academic journals in English. A total of 282 articles from 1256 authors were published in 176 journals in Chinese. The Chinese and English journals which published the most pulmonary nodules/early-stage lung cancer prediction model-related papers were Journal of Clinical Radiology and Frontiers in Oncology, respectively. Chest was the most frequently cited journal. China and the United States were the leading countries in the field of pulmonary nodules/early-stage lung cancer prediction models. The institutions represented by Fudan University had significant academic influence in the field. Analysis of keywords revealed that multi-omics, nomogram, machine learning and artificial intelligence were the current focus of research. Conclusion Over the last two decades, research on risk-prediction models for pulmonary nodules/early-stage lung cancer has attracted increasing attention. Prognosis, machine learning, artificial intelligence, nomogram, and multi-omics technologies are both current hotspots and future trends in this field. In the future, in-depth explorations using different omics should increase the sensitivity and accuracy of pulmonary nodules/early-stage lung cancer prediction models. More high-quality future studies should be conducted to validate the efficacy and safety of pulmonary nodules/early-stage lung cancer prediction models further and reduce the global burden of lung cancer.

Objective To investigate the occurrence status quo and influencing factors of low muscle mass(LMM)among young and middle-aged health examination population.Methods The young and middle-aged people undergoing the body composition analysis in this hospital from January to December 2023 were selected as the study subjects.The general data,body composition indices and biochemical indicators were col-lected.The body composition analysis was performed by the bioelectrical impedance analysis(BIA).LMM was diagnosed based on the skeletal muscle index.The univariate and multivariate logistic regression were used to analyze the influencing factors of LMN occurrence in the young and middle-aged health examination population.The receiver operating characteristic(ROC)curve and the area under the curve(AUC)were em-ployed to evaluate the predictive value of each indicator.Results A total of 2 351 people undergoing the phys-ical examination were included,aged 18-49 years old,366(15.57％)cases of LMM were detected out.The skeletal muscle index,sex,age,age group distribution,body mass index(BMI),body fat percentage(BFP),body fat percentage grade,visceral fat area(VFA),AST/ALT,Hb,serum creatinine,blood uric acid,HbA1c,fasting blood glucose,TC,LDL-C,HDL-C,TG and triglyceride-glucose index(TyG)had statistical differences between the LMM group and normal group(P＜0.05).Multivariate logistic regression revealed that the sex(OR=2.606,95％CI:1.755-3.870),BMI(OR=0.579,95％CI:0.538-0.623),BFP(OR=5.885,95％CI:4.176-8.292)and VFA(OR=0.955,95％CI:0.944-0.967)were the influencing factors for the LMM oc-currence in the young and middle-aged people undergoing the physical examination(P＜0.001).The ROC a-nalysis showed the AUC values of the sex,BMI,BFP and VFA for predicting LMM were 0.580,0.821,0.636 and 0.715 respectively,in which the predictive value of BMI was highest.Conclusion The population of fe-male,low BMI,high BFP and low VFA maybe the high-risk groups for LMM.The health management for the above-mentioned groups needs to be strengthened.

Gender recognition based on the analysis of fingerprint residue can assist investigators in narrowing down the scope of investigation and play an important role in the field of criminal investigation.This study established a quantitative analysis method for lipid substances in fingerprints based on gas chromatography-mass spectrometry(GC-MS).Fatty acids in fingerprints were methylated using sulfuric acid methanol derivatization reagent(7%,V/V),the extraction reagent was dichloromethane-methanol(1∶1,V/V)solution,the reaction temperature was 70℃and the heating time was 45 min.Quantitative analysis of the relative content of 23 kinds of fatty acids and squalene in fingerprints residue by different genders was conducted,and orthogonal partial least squares-discriminant analysis(OPLS-DA)was used to reduce the dimensionality of the quantitative results.A total of 13 kinds of components in the fingerprints were selected to maximize the difference in relative content between male and female fingerprints.Three machine learning models,including binary logistic regression(BLR),support vector machine(SVM)and random forest(RF),were further used as feature variables to classify the gender of fingerprints.The classification performance of each model was compared through five indicators,and it was found that the most suitable model for binary classification of fingerprint gender was SVM model.The results showed that the SVM fingerprint residual gender binary classification model established based on the relative content data of 13 kinds of lipid substances in fingerprints achieved a classification accuracy of 90%and an area under the receiver operating characteristic curve(AUC)value of 0.98.This study provided a new research method for detecting lipid components in fingerprints and a methodological basis for gender recognition of fingerprints.

Forensic medicine has been designated as a first-level discipline,presenting new opportunities and challenges for the development of forensic medicine.Since the 1980s,the establishment of foren-sic medicine discipline and the cultivation of high-level forensic talents have become hot topics in the development of forensic medicine in China.Since the 13th Five-Year Plan,the forensic team of Shanxi Medical University has been aiming at the forefront,proposing the development goals of"Five First-class"and the discipline development path"Six Major Achievements".It has selected benchmark disci-plines,identified gaps in disciplinary development,unified thoughts,formulated completion timelines,concentrated superior resources,assigned tasks to individuals,and created an"Organized Fill-in-the-Blank Format"forensic medicine discipline construction model with the characteristics of the new era.The construction model of forensic medicine has achieved good results in the goals,discipline frame-work,scientific research,talent cultivation,discipline team and platform construction,forming a rela-tively complete discipline construction and management system,and accumulating valuable experience for the construction of first-level discipline and high-level talent cultivation of forensic medicine.

ObjectivePrimary liver cancer, predominantly hepatocellular carcinoma (HCC), is a significant global health issue, ranking as the sixth most diagnosed cancer and the third leading cause of cancer-related mortality. Accurate and early diagnosis of HCC is crucial for effective treatment, as HCC and non-HCC malignancies like intrahepatic cholangiocarcinoma (ICC) exhibit different prognoses and treatment responses. Traditional diagnostic methods, including liver biopsy and contrast-enhanced ultrasound (CEUS), face limitations in applicability and objectivity. The primary objective of this study was to develop an advanced, light-weighted classification network capable of distinguishing HCC from other non-HCC malignancies by leveraging the automatic analysis of brightness changes in CEUS images. The ultimate goal was to create a user-friendly and cost-efficient computer-aided diagnostic tool that could assist radiologists in making more accurate and efficient clinical decisions. MethodsThis retrospective study encompassed a total of 161 patients, comprising 131 diagnosed with HCC and 30 with non-HCC malignancies. To achieve accurate tumor detection, the YOLOX network was employed to identify the region of interest (ROI) on both B-mode ultrasound and CEUS images. A custom-developed algorithm was then utilized to extract brightness change curves from the tumor and adjacent liver parenchyma regions within the CEUS images. These curves provided critical data for the subsequent analysis and classification process. To analyze the extracted brightness change curves and classify the malignancies, we developed and compared several models. These included one-dimensional convolutional neural networks (1D-ResNet, 1D-ConvNeXt, and 1D-CNN), as well as traditional machine-learning methods such as support vector machine (SVM), ensemble learning (EL), k-nearest neighbor (KNN), and decision tree (DT). The diagnostic performance of each method in distinguishing HCC from non-HCC malignancies was rigorously evaluated using four key metrics: area under the receiver operating characteristic (AUC), accuracy (ACC), sensitivity (SE), and specificity (SP). ResultsThe evaluation of the machine-learning methods revealed AUC values of 0.70 for SVM, 0.56 for ensemble learning, 0.63 for KNN, and 0.72 for the decision tree. These results indicated moderate to fair performance in classifying the malignancies based on the brightness change curves. In contrast, the deep learning models demonstrated significantly higher AUCs, with 1D-ResNet achieving an AUC of 0.72, 1D-ConvNeXt reaching 0.82, and 1D-CNN obtaining the highest AUC of 0.84. Moreover, under the five-fold cross-validation scheme, the 1D-CNN model outperformed other models in both accuracy and specificity. Specifically, it achieved accuracy improvements of 3.8% to 10.0% and specificity enhancements of 6.6% to 43.3% over competing approaches. The superior performance of the 1D-CNN model highlighted its potential as a powerful tool for accurate classification. ConclusionThe 1D-CNN model proved to be the most effective in differentiating HCC from non-HCC malignancies, surpassing both traditional machine-learning methods and other deep learning models. This study successfully developed a user-friendly and cost-efficient computer-aided diagnostic solution that would significantly enhances radiologists’ diagnostic capabilities. By improving the accuracy and efficiency of clinical decision-making, this tool has the potential to positively impact patient care and outcomes. Future work may focus on further refining the model and exploring its integration with multimodal ultrasound data to maximize its accuracy and applicability.

Objective To investigate the metabolic characteristics of patients with early-onset type 2 diabetes mellitus(T2DM)and to develop a risk prediction model for microvascular complications.Methods A retrospective study was conducted on 980 T2DM patients admitted for treatment between April 2020 and April 2024.Based on age at diagnosis,the patients were divided into two groups,an early-onset T2DM group(age at diagnosis<40 years,n=265)and a late-onset T2DM group(age at diagnosis≥40 years,n=715).Differences in metabolic indicators between the two groups were compared.Patients in the early-onset group were further divided into a complication subgroup(n=142)and a non-complication subgroup(n=123)based on the presence or absence of microvascular complications.Data on baseline characteristics,metabolic parameters,and laboratory indicators were collected and compared between the two groups.Multivariate logistic regression analysis was used to identify independent risk factors for microvascular complications,and a nomogram prediction model was constructed.The model's discriminative performance was assessed using receiver operating characteristic(ROC)curves,and its calibration was evaluated using calibration curves and the Hosmer-Lemeshow test.Decision curve analysis(DCA)was also performed to assess the model's clinical utility.Results Compared with the late-onset group,patients in the early-onset group exhibited more pronounced metabolic abnormalities,including higher body mass index(BMI),proportion of family history of diabetes mellitus,glycated hemoglobin(HbA1c)levels,total cholesterol(TC),triglycerides(TG),low-density lipoprotein cholesterol(LDL-C),triglyceride-glucose index(TyG),and lactate dehydrogenase(LDH)levels(all P<0.05),along with a shorter disease duration and lower levels of high-density lipoprotein cholesterol(HDL-C)(P<0.05).According to a multivariate analysis,systolic blood pressure(SBP),total bilirubin(TBIL),HDL-C,LDL-C,TyG,and LDH were identified as independent risk factors for microvascular complications in patients with early-onset T2DM.A predictive model based on these factors was established as the follows,Log(P)=-19.915+0.017×SBP-0.136×TBIL-1.241×HDL-C+0.684×LDL-C+0.769×TyG+0.050×LDH.The area under the ROC curve(AUC)was 0.864(95%CI,0.820-0.907),and the Hosmer-Lemeshow test indicated good model fit(χ2=10.286,P=0.246).The slope of the DCA curve was also close to 1.Conclusion The nomogram prediction model based on SBP,TBIL,HDL-C,LDL-C,TyG,and LDH demonstrates good predictive performance for microvascular complications and can provide a reference for clinical risk stratification and individualized intervention.