Objective To evaluate the predictive value of different machine learning models constructed in a multicenter environment for potential organ donors and verify their clinical application feasibility. Methods The study included 2 000 inpatients admitted to five domestic tertiary hospitals from January 2020 to December 2023, who met the criteria for potential organ donation assessment. They were randomly divided into a training set and an internal validation set (7∶3). Another 300 similar patients admitted to the First Affiliated Hospital of Harbin Medical University from January 2024 to April 2025 were included as an external validation set. The area under the curve (AUC), sensitivity, specificity, accuracy and F1-score of three models were compared, and the consistency of the potential organ donor determination process was tested. Multivariate logistic regression analysis was used to identify predictive factors of potential organ donors. Decision curve analysis (DCA) was employed to verify the resource efficiency of each model, and the threshold interval and intervention balance point were assessed. Results Apart from age, there were no significant differences in other basic characteristics among the centers (all P>0.05). The consistency of the potential organ donor determination process among researchers in each center was good [all 95% confidence interval (CI) lower limits >0]. In the internal validation set, the XGBoost model had the best predictive performance (AUC=0.92, 95% CI 0.89-0.94) and the best calibration (P=0.441, Brier score 0.099). In the external validation set, the XGBoost model also had the best predictive performance (AUC=0.91, 95% CI 0.88-0.94), outperforming logistic regression and random forest models. Multivariate logistic regression showed that mechanical ventilation had the greatest impact (odds ratio=2.06, 95% CI 1.54-2.76, P<0.001). DCA indicated that the XGBoost model had the highest net benefit in the threshold interval of 0.2-0.6. The “treat all” strategy only had a slight advantage at extremely low thresholds. The recommended threshold interval, which balances intervention costs and clinical benefits, considers ≥50% positive predictive value (PPV) and ≤50 referrals per 100 high-risk patients. Conclusions The XGBoost model established in a multicenter environment is accurate and well-calibrated in predicting potential organ donors. Combined with DCA, it may effectively guide the timing of clinical interventions and resource allocation, providing new ideas for the assessment and management of organ donation after brain death.

Background: Visceral white adipose tissue (vWAT) inflammation is a critical pathology of obesity-caused heart damage and is closely associated with adipocyte endoplasmic reticulum (ER) dysfunction. Serine (or cysteine) peptidase inhibitor, clade A, member 3C (Serpina3c) has been identified as an adipokine with anti-vWAT inflammatory effects. However, it remains unclear whether Serpina3c deficiency promotion of vWAT inflammation involves adipocyte ER dysfunction and whether it further contributes to heart damage in obesity. Methods: Wild type and Serpina3c knockout (Serpina3c–/–) mice were fed a high-fat diet (HFD) for 12 weeks. An adeno-associated virus (AAV) was injected locally into epididymal white adipose tissue (eWAT) of Serpina3c–/– mice to induce eWAT-adipocyte- specific overexpression of Serpina3c (AAV-Serpina3c) or knockdown of hypoxia-inducible factor 1α (AAV-shHIF1α). In vitro experiments were performed in 3T3-L1 adipocytes. Results: Serpina3c–/– mice exhibited more severe eWAT, serum and heart inflammation after HFD feeding. Consistently, these adverse phenotypes were mitigated in AAV-Serpina3c and AAV-shHIF1α mice. Mechanistically, ER oxidoreductase 1α (Ero1α) and protein disulfide isomerase (PDI) family members PDIA3 and PDIA4 were found to be target genes of HIF1α. In the obese mice, Serpina3c deficiency caused adipocyte more hypertrophy, and activated HIF1α-Ero1α/PDI mediated ER overoxidation and ER stress in eWAT. Subsequently, this led to increased adipocyte apoptosis and chemokine production and decreased adiponectin expression, which promoted macrophage infiltration and M1 polarization in eWAT, thus exacerbating eWAT inflammation and ultimately facilitating serum and distal heart inflammation. Conclusion These findings indicate that Serpina3c is a significant regulator of adipocyte ER redox homeostasis, thus highlighting Serpina3c as a potential therapeutic target for obesity-related eWAT inflammation and heart damage.

Electroencephalography (EEG) is a non-invasive neurophysiological monitoring technique. It records the electrical activity of the cerebral cortex using electrodes placed on the scalp surface. Owing to its high safety, portability, and millisecond-level temporal resolution, EEG has been widely utilized in a variety of fields, including clinical diagnosis, brain-computer interfaces (BCIs), and cognitive neuroscience research. However, due to its microvolt-level amplitude, EEG is highly susceptible to various artifacts, including electrooculographic (EOG), electrocardiographic (ECG), electromyographic (EMG), and power line interference (PLI). These artifacts can obscure genuine neural activity and introduce spurious electrophysiological features. Consequently, they may compromise EEG signal quality, thereby reducing the reliability of downstream analyses. To address this issue, numerous EEG artifact removal methods have been developed, including both traditional denoising techniques and deep learning-based approaches. Traditional EEG denoising methods have long served as the primary solutions for artifact removal. Representative approaches include filtering, regression, and blind source separation. Although these methods have demonstrated effectiveness in specific scenarios, they suffer from several inherent limitations. Filtering assumes that artifacts and EEG signals can be separated in the frequency domain, but many artifacts, such as EOG and EMG, overlap with EEG spectra, which may lead to the loss of valuable neural information. Regression methods require high-quality artifact references to estimate and subtract contaminations, limiting their effectiveness in reference-free scenarios. Blind source separation can remove artifacts without external references, but it typically requires the number of EEG channels to exceed the number of sources, restricting its application in single- or low-channel EEG recordings. Deep learning-based EEG denoising methods address these limitations effectively. First, they learn the nonlinear mapping between contaminated and clean EEG directly from data in an end-to-end manner. This approach does not rely on assumptions about spectral separability, thereby preserving neural activity more completely. Second, the reference information is incorporated during the training phase, allowing the trained model to perform artifact removal independently without external references. Third, deep learning models can be flexibly designed to accommodate various recording setups, achieving robust denoising for both high-density and single-channel EEG. Collectively, these advantages enable deep learning-based methods to overcome the main challenges of traditional approaches, providing more accurate and reliable EEG signal recovery. The superior denoising performance of deep learning-based EEG denoising methods has attracted increasing attention in EEG artifact removal research. As a result, many deep learning-based denoising methods have been developed and successfully applied in neural engineering areas. However, a systematic review of the techniques and applications in this field is still lacking. To address this gap, this paper reviews recent advances in deep learning-based EEG denoising from four perspectives: technical principle, benchmark dataset, denoising model, and evaluation method. Representative applications in neural signal analysis and BCI decoding are also summarized. Furthermore, the advantage, existing challenge, and future research direction of deep learning-based EEG denoising are discussed. This review aims to provide valuable theoretical insights and technical guidance for researchers. It is also expected to promote further advances and broader applications of deep learning-based EEG denoising techniques.

Objective To explore the mediating role of activities of daily living (ADL) in pain and depressive symptoms in the elderly in China. Methods Utilizing the data from 2020 China Health and Retirement Longitudinal Study, 4403 Chinese elderly individuals aged ≥ 60 years old were selected as the research subjects. Depression Scale (CES-D 10) of the Center for Epidemiological Survey and ADL scale were used in the study. The PROCESS4.1 macro was used to test the mediating effect of daily living activities between pain and depressive symptoms, and the Bootstrap method was applied for verification of the mediating variables. Results A total of 2368 cases of depressive symptoms were detected in the elderly in China, with a detection rate of 53.78%. Pain was positively correlated with depressive symptoms (r=0.27, P<0.01), and activities of daily living were negatively correlated with pain and depressive symptoms (r=-0.27, -0.337, P<0.01). The results showed that the total effect value of pain on depressive symptoms was 0.33, the direct effect value was 0.24, and the mediating effect value of daily living activities was 0.09, accounting for 27.27%. Conclusion Pain and activities of daily living are important factors influencing depressive symptoms in the elderly, and activities of daily living play a partial mediating role in the relationship between pain and depressive symptoms in the elderly.

ObjectiveTo investigate the value of Fried Frailty Phenotype （FFP）， liver frailty index （LFI）， and Short Physical Performance Battery （SPPB） in predicting 2-year all-cause mortality and decompensation events in patients with liver cirrhosis. MethodsA total of 277 patients with liver cirrhosis who were hospitalized in Beijing Friendship Hospital， Capital Medical University， from December 2020 to December 2021 were enrolled， and FFP， LFI， and SPPB were used to assess the state of frailty. Based on the scores of each tool， these patients were divided into frail and non-frail groups. These three tools were compared in terms of consistency and independent predictive performance. The primary endpoints were 2-year all-cause mortality rate and composite endpoints （death+decompensation events）， and the Cox regression analysis， the receiver operating characteristic （ROC） curve， net reclassification index （NRI）， and integrated discrimination improvement （IDI） index were used to analyze the predictive value of the three tools. Normally distributed continuous data were compared between two groups using the independent samples t-test， while non-normally distributed continuous data were compared using the Mann-Whitney U test. Categorical data were compared between groups using the chi-square test or Fisher’s exact test. The agreement among different frailty tools was evaluated using Cohen’s Kappa statistic. The Kaplan-Meier survival curve was plotted， and a survival analysis was performed using the log-rank test. ResultsThe prevalence rate of frailty assessed by FFP， LFI， and SPPB was 37.2%， 22.4%， and 20.2%， respectively， with a moderate consistency between FFP and LFI/SPPB （κ=0.57， 95% confidence interval ［CI］： 0.47 — 0.67； κ=0.51， 95%CI： 0.41 — 0.62） and a relatively high consistency between LFI and SPPB （κ=0.87， 95%CI： 0.80 — 0.94）. Compared with the non-frailty group， the frailty group had significantly higher all-cause mortality rate and incidence rate of composite endpoints （P0.001）. After multivariate adjustment， FFP， LFI， and SPPB had a hazard ratio of 2.42（95%CI： 1.51 — 5.11）， 2.21（95%CI： 1.11 — 4.42）， and 2.21（95%CI： 1.14 — 4.30）， respectively， in predicting all-cause mortality， as well as a hazard ratio of 2.51（95%CI： 1.61 — 3.91）， 2.40（95%CI： 1.51 — 3.80）， and 2.20（95%CI： 1.39 — 3.47）， respectively， in predicting composite endpoints. Compared with Child-Pugh score， FFP had a significantly greater area under the ROC curve （AUC） in predicting all-cause mortality （0.79 vs 0.69， P=0.032） and composite endpoints （0.75 vs 0.68， P=0.044）. Frailty assessment tools combined with Child-Pugh score significantly improved the performance in predicting all-cause mortality and composite endpoints， with an AUC of 0.81 — 0.82 and 0.77 — 0.78， respectively （P0.05）. NRI and IDI analyses further confirmed the improvement of the combined model in classification （all P0.001）. ConclusionFFP， LFI， and SPPB can independently predict adverse outcomes in patients with liver cirrhosis， among which FFP has the best predictive performance， and the combination of frailty assessment tools with Child-Pugh score can significantly enhance the accuracy of prognostic evaluation.

As a hot spot in clinical research today, immune checkpoint inhibitor has been recommended by guidelines in the first- and second-line treatments of advanced cervical cancer as immune monotherapy or combination therapy. It has also achieved good efficacy in clinical practice. In locally advanced cervical cancer, immune checkpoint inhibitors have been included in the guidelines for adjuvant therapy, and good tumor regression effects have been achieved in clinical practice. Based on the results of existing trials, immune checkpoint inhibitors have also shown good clinical potential as neoadjuvant therapy. Furthermore, the issue of immunotherapy rechallenge has increasingly captured clinicians’ attention, offering a potential new therapeutic strategy for cervical cancer patients with prior immunotherapy exposure. In this article, the clinical application and research progress of immune checkpoint inhibitors in the treatment of cervical cancer in recent years are summarized to provide valuable ideas and directions for clinical treatment.

Objective To analyze the characteristics and risk factors associated with postoperative deep vein thrombosis(DVT)of the lower extremities in patients undergoing surgery for lumbar degenerative diseases.Methods A retrospective analysis was conducted on clinical data from 298 patients who were hospitalized for lumbar degenerative diseases and underwent lumbar spine surgery treatment in the First Affiliated Hospital of Nanchang University from October 1,2022 to April 15,2023.Patients were divided into DVT group(n=71)and non-DVT group(n=227)according to whether DVT of the lower limbs occurred within 1 week postoperatively.The incidence and distribution characteristics of postoperative DVT were analyzed.Univariate and binary logistic regression analyses were performed to identify risk factors for DVT,and receiver operating characteristic(ROC)curves were used to determine cut-off values for relevant risk factors.Results A total of 298 patients were included,among whom 159 were males(53.4%)and 139 were females(46.6%),with an average age of(64.5±9.8)years.DVT occurred in 71 patients,and the incidence of lower extremity DVT was 23.8%.In the DVT group,there were 49 cases(69.0%)of intermuscular vein thrombosis,and 22 cases of other types of thrombosis(7 cases of peroneal vein thrombosis,4 cases of posterior tibial vein thrombosis,3 cases of common femoral vein thrombosis,1 case of anterior tibial vein thrombosis,and 7 cases of multiple thrombosis);58 cases(81.7%)had DVT in one lower extremity,and 13 cases(18.3%)had DVT in both lower extremities.Univariate analysis results showed that age,body mass index(BMI),length of hospital stay,history of hypertension,operative time,and intraoperative blood loss were associated with the occurrence of lower extremity DVT after surgery for lumbar degenerative diseases(P<0.05).Binary logistic regression analysis results indicated that older age(OR=1.079,P<0.01),higher BMI(OR=1.130,P=0.01),history of hypertension(OR=2.992,P<0.01),and larger intraoperative blood loss(OR=1.002,P=0.03)were independent risk factors for the occurrence of lower extremity DVT.ROC curve analysis demonstrated that patients with age>58.5 years,BMI>24.01 kg/m2,history of hypertension,and intraoperative blood loss>550 ml had a significantly increased risk of postoperative lower limb DVT.Conclusions The incidence of lower extremity DVT after surgery for lumbar degenerative disease is high,and intermuscular venous thrombosis is more common.Older age,higher BMI,history of hypertension,and larger intraoperative blood loss are independent risk factors for the occurrence of lower extremity DVT after surgery.

Objective To explore the status of nursing work engagement, nursing work environment and emotional intelligence and their relationship among nurses who work in hematopoietic stem cell transplantation (HSCT) units. Methods A total of 225 HSCT nurses were selected as study subjects by convenience sampling method. Utrecht Work Engagement Scale, Nursing Work Environment Scale and Emotional Intelligence Scale were used to assess the work engagement, work environment and emotional intelligence among these nurses. AMOS 23.0 software was used to construct the structural equation model. Results The median and 25th and 75th percentiles of the score of work engagement of the research subjects were 59.0 (54.0, 64.0) points. The average scores of the nursing work environment and emotional intelligence were (117.8±21.5) and (58.8±10.7) points, respectively. The score of work engagement was positively correlated with the scores of the nursing work environment and emotional intelligence (rank correlation coefficients were 0.550 and 0.431, respectively, both P<0.01). The total score of the nursing work environment was positively correlated with the total score of emotional intelligence (correlation coefficient was 0.271, P<0.01). The nursing work environment influenced the work engagement status of HSCT nurses through the mediating effect of emotional intelligence, with an indirect effect of 0.115 (95% confidence interval: 0.201-0.365), accounting for 20.4% of the total effect. Conclusion Emotional intelligence is a mediating variable between the nursing work environment and work engagement of HSCT nurses.

Liposomes serve as critical carriers for drugs and vaccines,with their biological effects influenced by their size.The microfluidic method,renowned for its precise control,reproducibility,and scalability,has been widely employed for liposome preparation.Although some studies have explored factors affecting liposomal size in microfluidic processes,most focus on small-sized liposomes,predominantly through experimental data analysis.However,the production of larger liposomes,which are equally significant,remains underexplored.In this work,we thoroughly investigate multiple variables influencing liposome size during microfluidic preparation and develop a machine learning(ML)model capable of accurately predicting liposomal size.Experimental validation was conducted using a staggered herringbone micromixer(SHM)chip.Our findings reveal that most investigated variables significantly influence liposomal size,often interrelating in complex ways.We evaluated the predictive performance of several widely-used ML algorithms,including ensemble methods,through cross-validation(CV)for both lipo-some size and polydispersity index(PDI).A standalone dataset was experimentally validated to assess the accuracy of the ML predictions,with results indicating that ensemble algorithms provided the most reliable predictions.Specifically,gradient boosting was selected for size prediction,while random forest was employed for PDI prediction.We successfully produced uniform large(600 nm)and small(100 nm)liposomes using the optimised experimental conditions derived from the ML models.In conclusion,this study presents a robust methodology that enables precise control over liposome size distribution,of-fering valuable insights for medicinal research applications.

OBJECTIVES: To investigate the protective effects of dexmedetomidine (DEX) against heat stress (HS)-induced oncosis in human skeletal muscle cells (HSKMCs) and its underlying mechanisms. METHODS: A HSKMC model of HS-induced oncosis were established by 43 ℃ water bath for 4 h, and the effects of treatments with 30 μmol/L DEX, ML385 (a Nrf2 inhibitor) +DEX, si-Nrf2+HS, and si-Nrf2+DEX prior to modeling on cell viability was assessed using CCK-8 assay. Oncosis characteristics were evaluated using transmission electron microscopy and Annexin V-FITC/PI flow cytometry. The oxidative stress markers (GSH, GSH-Px, MDA, SOD and ROS), mitochondrial membrane potential, energy metabolism, and inflammatory cytokines (TNF-α, IL-6 and IL-1β) in the cells were quantified using standard kits, and the expressions of porimin, caspase-3 and Nrf2 pathway proteins were analyzed using Western blotting and qRT-PCR. RESULTS: HS induced typical oncotic features in HSKMCs including organelle swelling and cytoplasmic vacuolization. DEX pretreatment significantly attenuated these changes, reduced Annexin V⁺/PI⁺ cell ratio and cellular porimin expression, and lowered the levels of ROS and MDA while restoring GSH and SOD levels. DEX pretreatment also significantly increased the mitochondrial membrane potential and ATP level, upregulated the expressions of Nrf2, p-Nrf2, HO-1 and NQO1, and suppressed the expressions of TNF-α, IL-6 and IL-1β. The protective effects of DEX were obviously attenuated by interventions with ML385 or si-Nrf2. CONCLUSIONS DEX mitigates HS-induced HSKMC oncosis by activating the Nrf2/HO-1 pathway to relieve oxidative stress, mitochondrial dysfunction, and inflammatory responses.
Humans ; Dexmedetomidine/pharmacology* ; NF-E2-Related Factor 2/metabolism* ; Oxidative Stress/drug effects* ; Heat-Shock Response/drug effects* ; Signal Transduction/drug effects* ; Membrane Potential, Mitochondrial ; Muscle, Skeletal/cytology* ; Heme Oxygenase-1/metabolism* ; Apoptosis/drug effects*