Wuxi Third People's Hospital

Bladder cancer is one of the most prevalent malignancies of the urinary system and is associated with high morbidity and mortality. With advances in medical image analysis, deep learning has shown promise for automated bladder cancer classification using magnetic resonance imaging (MRI). However, clinical deployment remains challenging due to substantial inter-center distributional discrepancies and limited feature discriminability between non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC). To address these challenges, we propose a Domain-Adaptive Deep Contrastive Network (DADCNet) for MRI-based bladder cancer classification. The proposed framework jointly incorporates source- and target-domain samples during feature learning to obtain domain-invariant yet discriminative representations, thereby improving cross-center generalization. In addition, a deep contrastive learning strategy is introduced to enhance inter-class separability and intra-class compactness, leading to more robust classification. Experiments conducted on a multi-center bladder cancer MRI dataset demonstrate that DADCNet consistently outperforms existing convolutional neural network- and Transformer-based methods, achieving an accuracy of 0.955, an F1-score of 0.955, and an area under the curve of 0.991.

HIV infection among older adults in rural China has risen significantly in recent years, yet epidemiological characteristics and intervention effectiveness remain understudied. We described the characteristics of 179 HIV/AIDS patients aged ≥ 50 years in rural Nantong, China, then conducted a community-based 1:1 case-control study of 179 HIV/AIDS patients aged ≥ 50 years and developed a predictive nomogram. Mediation analysis was applied to explore the role of HIV knowledge, and the effectiveness of a comprehensive community intervention was assessed among 1206 community individuals. There were 179 elderly HIV/AIDS participants recruited, 55.9% of them were homosexual transmission, and 84.4% were identified through passive pathways. The case-control showed that the protective factors associated with HIV infection included higher Perceived Social Support Score (PSSS) (OR = 0.023) and knowledge awareness (OR = 0.055), while the risk factors included being unemployed (OR = 4.852), having extramarital sex (OR = 30.264), and the tolerant attitude to extramarital sex (OR = 7.124). The C-index is 0.9 and the Dxy is 0.913 with Spiegelhalter Z-test, indicating that the nomogram has good predictive accuracy. Mediation analysis indicated that knowledge mediated the impact of PSSS, extramarital sex, and attitude towards extramarital sex on HIV infection, accounting for 30.8%, 26.6% and 14.1%of the effects, respectively. Six months after implementation of the comprehensive intervention, there was a significant increase in HIV knowledge awareness (from 41.0% to 84.4%) and a decrease in extramarital sex(from 2.4% to 0.8%). Appropriate interventions for the rural elderly with low education can also effectively reduce AIDS related risky behaviors. PSSS can be considered as a potential screening factor to predict the risk of HIV infection.

This study aimed to identify risk factors associated with unsuccessful treatment outcomes among newly diagnosed drug-sensitive pulmonary tuberculosis (PTB) patients in rural eastern China and to develop a prediction model for treatment outcomes. This study analyzed 838 newly diagnosed drug-sensitive PTB patients in rural eastern China (2021-2023). Treatment outcomes (unsuccessful treatment) were assessed using WHO guidelines. The cohort was randomly divided into a training set (70%) and a validation set (30%) for internal validation. Multivariate logistic regression identified predictors, including age, malnutrition, comorbidities, hemoglobin levels, and sputum smear grades. Decision curve analysis (DCA) was performed to evaluate the clinical utility of the prediction model by quantifying the net benefit across a range of threshold probabilities. The prediction model identified six independent predictors of unsuccessful treatment outcomes: diabetes, chronic lung disease, alcohol use, hypoalbuminemia, anemia, and sputum smear grades. The area under the receiver operating characteristic curve (AUC) was 0.754 (95% CI: 0.676-0.833), indicating good discriminative ability. The model demonstrated moderate accuracy across three risk categories. A nomogram was developed to visually represent the model, enabling clinicians to estimate individual patient risk based on these six predictors. Additionally, an online calculator was created to facilitate easy and practical application of the model in clinical settings. Decision curve analysis (DCA) further validated the clinical utility of the model, showing a significant net benefit across a wide range of threshold probabilities (2-54%), supporting its applicability for guiding clinical decision-making. The prediction model serves as a valuable tool for clinicians to identify high-risk PTB patients and tailor interventions effectively. This approach can enhance treatment strategies and contribute to better TB control in rural eastern China.

BackgroundThere is a disease spectrum of ischemic stroke (IS) and obstructive sleep apnea (OSA), which are often comorbid in the same patient and consequently increases prognostic risk. Continuous positive airway pressure (CPAP) therapy is the primary treatment for stroke-related OSA; however, its positive effects on patient prognosis and the underlying mechanisms remain controversial.ObjectiveThis study aimed to investigate the impact of CPAP therapy on the recovery of IS patients with moderate-to-severe OSA and to identify biomarkers significantly associated with prognosis to assess their predictive value for short-term neurological outcomes. The findings are expected to optimize treatment strategies and improve overall patient outcomes.MethodsA total of 141 patients with IS combined with moderate-to-severe OSA admitted to the Affiliated Hospital of Yangzhou University from October 2022 to August 2024 were enrolled. Patients were divided into a CPAP group (n = 68) and a control group (n = 73). Both groups received systematic treatment and were followed up until 1 month after the onset of stroke symptoms. The CPAP group initiated therapy within 48 h of stroke onset (ResMed AutoCPAP, pressure 4–20 cmH2O) for 14 days (adherence criterion: ≥4 h/day). Baseline data, sleep and stroke-related questionnaires, polysomnography (PSG) parameters, and sleep spindle characteristics were collected. Neurological functional outcomes were reassessed at the end of the follow-up period, and differences between the two groups were analyzed. Prognostic factors were identified using Spearman correlation analysis and ordered logistic regression.ResultsCompared with those in the control group, patients in the CPAP group had lower modified Rankin scale (mRS) and National Institutes of Health Stroke Scale (NIHSS) scores after treatment (p < 0.05), while the Barthel index (BI) did not significantly differ. Spearman correlation analysis revealed that mRS scores were positively correlated with the apnea-hypopnea index (AHI), the AHI during the rapid eye movement stage (REM-AHI) and the AHI during the non-rapid eye movement stage (NREM-AHI) (all p < 0.05) and negatively correlated with the non-rapid eye movement stage 2 sleep spindle density (N2-SSD), the non-rapid eye movement stage 3 (N3) sleep percentage, and the mean pulse oxygen saturation (Mean SpO2) (all p < 0.05). Logistic regression revealed that N2-SSD, Mean SpO2, and REM-AHI were significant predictors of mRS scores (all p < 0.05).ConclusionCPAP therapy enhances sleep microstructure and oxygenation parameters, which improves sleep quality. N2-SSD, REM-AHI, and Mean SpO2 are mechanistically linked to functional prognosis and CPAP exerts therapeutic effects through the modulation of these biomarkers. Early CPAP intervention targeting REM-AHI and N2-SSD demonstrates prognostic benefits, which suggests that sleep microstructure-specific metrics may serve as precision therapeutic targets.

Abstract Metabolic dysfunction‐associated steatotic liver disease (MASLD) has become an important liver disease worldwide due to the rising rate of type 2 diabetes, obesity, and other metabolic disorders. The early detection of MASLD is vital to slow progression to cirrhosis and hepatocellular carcinoma; yet, existing research tools for studying hepatic steatosis are limited in terms of their effectiveness and invasiveness. Herein, a near‐infrared (NIR) fluorescent probe, TCF‐FBN, for the selective imaging of lipid droplets (LDs) in MASLD is reported. TCF‐FBN exhibits bright NIR fluorescence and a large Stokes shift in nonpolar solvents, and minimal fluorescence in polar solvents, making it suitable for biological imaging. The probe is bound to bovine serum albumin (BSA) through supramolecular forces to enhance its biocompatibility, and the resulting TCF‐FBN@BSA ensemble is used for fluorescence imaging of LDs generated in hepatocytes. Notably, the modification of TCF‐FBN@BSA with galactosyl ligands producing TCF‐FBN@Gal‐BSA enables the targeted imaging of the liver of mice with MASLD via asialoglycoprotein receptor (ASGPR)‐mediated biorecognition.

Abstract Objective To propose a deep learning model and explore its performance in the auxiliary diagnosis of lung cancer associated with cystic airspaces (LCCA) in computed tomography (CT) images. Methods This study is a retrospective analysis that incorporated a total of 342 CT images, comprising 272 images from patients diagnosed with LCCA and 70 images from patients with pulmonary bulla. A deep learning model named LungSSFNet, developed based on nnUnet, was utilized for image recognition and segmentation by experienced thoracic surgeons. The dataset was divided into a training set (245 images), a validation set (62 images), and a test set (35 images). The performance of LungSSFNet was compared with other models such as UNet, M2Snet, TANet, MADGNet, and nnUnet to evaluate its effectiveness in recognizing and segmenting LCCA and pulmonary bulla. Results LungSSFNet achieved an intersection over union of 81.05% and a Dice similarity coefficient of 75.15% for LCCA, and 93.03% and 92.04% for pulmonary bulla, respectively. These outcomes demonstrate that LungSSFNet outperformed many existing models in segmentation tasks. Additionally, it attained an accuracy of 96.77%, a precision of 100%, and a sensitivity of 96.15%. Conclusion LungSSFNet, a new deep-learning model, substantially improved the diagnosis of early-stage LCCA and is potentially valuable for auxiliary clinical decision-making. Our LungSSFNet code is available at https://github.com/zx0412/LungSSFNet.

BackgroundMicrovascular invasion (MVI) is an important risk factor for early postoperative recurrence of hepatocellular carcinoma (HCC). Based on gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) images, we developed a novel radiomics model. It combined bi-regional features and two machine learning algorithms. The aim of this study was to validate its potential value for preoperative prediction of MVI.MethodsThis retrospective study included 304 HCC patients (training cohort, 216 patients; testing cohort, 88 patients) from three hospitals. Intratumoral and peritumoral volumes of interest were delineated in arterial phase, portal venous phase, and hepatobiliary phase images. Conventional radiomics (CR) and deep learning radiomics (DLR) features were extracted based on FeAture Explorer software and the 3D ResNet-18 extractor, respectively. Clinical variables were selected using univariate and multivariate analyses. Clinical, CR, DLR, CR-DLR, and clinical-radiomics (Clin-R) models were built using support vector machines. The predictive capacity of the models was assessed by the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity.ResultsThe bi-regional CR-DLR model showed more gains and gave better predictive performance than the single-regional models or single-machine learning models. Its AUC, accuracy, sensitivity, and specificity were 0.844, 76.9%, 87.8%, and 69.1% in the training cohort and 0.740, 73.9%, 50%, and 84.5% in the testing cohort. Alpha-fetoprotein (odds ratio was 0.32) and maximum tumor diameter (odds ratio was 1.270) were independent predictors. The AUCs of the clinical model and the Clin-R model were 0.655 and 0.672, respectively. There was no significant difference in the AUCs between all the models (P > 0.005).ConclusionBased on Gd-EOB-DTPA-enhanced MRI images, we focused on developing a radiomics model that combines bi-regional features and two machine learning algorithms (CR and DLR). The application of the new model will provide a more accurate and non-invasive diagnostic solution for medical imaging. It will provide valuable information for clinical personalized treatment, thereby improving patient prognosis.Clinical trial numberNot applicable.

The microRNA-200 family plays a key role in inflammatory and vascular processes, making it a relevant target for Kawasaki disease, a vasculitis with coronary complications. This study aimed to evaluate the diagnostic potential of urinary exosomal microRNA-200 family members in Kawasaki disease patients. Urine samples from 15 Kawasaki disease patients and 15 healthy controls underwent total exosome isolation and high-throughput sequencing. Differential expression of microRNA-200 family members was validated using quantitative real-time polymerase chain reaction. Diagnostic potential was assessed via receiver operating characteristic analysis, and correlations with clinical parameters were evaluated using Spearman correlation. High-throughput sequencing identified upregulation of microRNA-429, microRNA-200b-3p/5p, microRNA-141-3p, microRNA-200a-3p/5p, and microRNA-200c-3p in Kawasaki disease patients. We confirmed significant upregulation of microRNA-200a-3p/5p, microRNA-200b-3p/5p, and microRNA-429, with receiver operating characteristic analysis showing high diagnostic potential for these microRNAs (area under the curves of 0.844, 0.791, 0.942, 0.842, and 0.898, respectively) and a combined analysis yielding a perfect area under the curve of 1.000. MicroRNA-141 and microRNA-200c-3p/5p, however, showed no significant diagnostic value. MicroRNA-200a-3p and microRNA-200a-5p were positively correlated with white blood cells, platelet counts, and C-reactive protein, while microRNA-200b-3p and microRNA-429 were positively correlated with white blood cells, platelet counts, erythrocyte sedimentation rate, and C-reactive protein. microRNA-200b-5p showed moderate correlations with platelet counts and erythrocyte sedimentation rate. Urinary exosomal microRNA-200 family members, especially microRNA-200a-3p/5p, microRNA-200b-3p/5p, and microRNA-429, demonstrate strong diagnostic potential for Kawasaki disease, correlating with key inflammatory markers. MicroRNA-141 and microRNA-200c did not demonstrate significant diagnostic utility.