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.

Background: The surgical neck of the humerus is located approximately 2–3 cm distal to the anatomical neck (collum anatomicum) at the junction of the greater and lesser tubercles and the humeral shaft. Owing to the transition between cortical and cancellous bones, this region is biomechanically vulnerable and prone to fractures, particularly in older adults. Fractures of the surgical neck of the humerus are classified under the International Classification of Diseases, 10th Revision (ICD-10), with diagnostic codes S42.21.3, S42.22, and S42.23. National statistics indicate that 210,763 new trauma cases were reported in Mongolia in 2023, representing an 11.6% increase compared to the previous year, of which 11,452 cases (5.4%) involved shoulder and humeral injuries. Aim: This study aimed to compare the clinical effects of traditional Mongolian bone-setting therapy and surgical treatment on pain intensity, shoulder function, range of motion, and activities of daily living in patients with surgical neck of the humerus fractures. Materials and Methods: Based on the Department of Traditional Therapies of the Mongolian National University of Medical Sciences (MNUMS) and the Department of Orthopedic Trauma of the Inner Mongolia International Hospital, Inner Mongolia Autonomous Region, China, a total of 60 patients diagnosed with surgical neck fractures of the humerus were enrolled in this study between January 2024 and December 2025. The patients were randomly assigned into two groups: a Mongolian traditional bone-setting therapy group (n=30) and a surgical treatment group (n=30). Assessments were conducted before treatment and at 1 month and 3 months after treatment. Shoulder joint range of motion, activities of daily living, and muscle strength were evaluated using the Constant–Murley score, while pain intensity was assessed using the Visual Analog Scale (VAS). The collected data were systematically analyzed, and the therapeutic outcomes of the two treatment modalities were comprehensively compared and summarized. Result: The baseline demographic and clinical characteristics were comparable between the groups. In the traditional bone-setting group, VAS scores decreased significantly, whereas Constant–Murley scores improved markedly at both follow-up points. The overall outcomes in the traditional bone-setting group were significantly better than those in the surgical group (P<0.05). Conclusion Traditional Mongolian bone-setting therapy offers rapid pain relief, superior long-term functional recovery, and fewer complications, making it a valuable treatment option, particularly for elderly patients and those unsuitable for surgery.

Aim To explore the specific mechanism by which scutellarin(Scu)antagonizes the injury of human aortic endothelial cells(HAEC)induced by 2,2-azobis(2-methylpropylimidate)dihydrochloride(AAPH)by regulating the protein kinase RNA-like endoplasmic reticulum kinase(PERK)-nuclear factor erythroid 2-related factor 2(NRF2)/ac-tivating transcription factor 4(ATF4)-C/EBP homology protein(CHOP)pathway.Methods HAEC were pre-pro-tected by Scu and then injured by AAPH to explore the molecular mechanism of Scu on HAEC injury.The cells were di-vided into control group,AAPH group,AAPH+Scu low,medium and high groups.The contents of superoxide dismutase(SOD),malondialdehyde(MDA),glutathione peroxidase(GSH-Px)and glutathione S-transferase(GSH-ST)in the cells were measured.The content of reactive oxygen species(ROS)in cells was detected by fluorescent probe,and the apop-tosis rate was detected by Annexin V-FITC/PI method.The mRNA expression of PERK and eIF2α in cells was detected by RT-qPCR.The protein expression of glucose regulated protein 78(GRP78),PERK,p-PERK,eukaryotic initiation factor 2α(eIF2α),p-eIF2α,ATF4,CHOP,Nrf2,Bcl-2,p53 up-regulated modulator of apoptosis(PUMA),Caspase-3 and Caspase-12 in cells was detected by Western blot.In order to further study the molecular mechanism of Scu against HAEC injury,gene silencing technology was used to inhibit the expression of PERK in HAEC.The cells were divided in-to five groups:control group,AAPH+si-con group,AAPH+Scu+si-con group,AAPH+si-PERK group,AAPH+si-PERK+Scu group.The mRNA expression of PERK and eIF2α in cells after si-PERK interference was detected by RT-qPCR.The protein expression of PERK,p-eIF2α,eIF2α,ATF4,CHOP,Nrf2,Bcl-2,PUMA,Caspase-3 and Caspase-12 in cells after si-PERK interference was detected by Western blot.Results The content of ROS and the rate of apoptosis were significantly reduced after Scu intervention(P＜0.01).Scu could down-regulate the mRNA expression of PERK and eIF2α,and down-regulate the protein expression of GRP78,p-PERK,p-eIF2α,ATF4,CHOP,PUMA,Caspase-3,Caspase-12 and up-regulate the protein expression of Nrf2 and Bcl-2(P＜0.01).After interference with si-PERK,there were significant differences in the protein expression of PERK,p-eIF2α,ATF4,CHOP,Nrf2,Bcl-2,PUMA,Caspase-3,Caspase-12,as well as the mRNA expression of PERK and eIF2α in cells compared to before interference(P＜0.01).It is proved that Scu could anti-endoplasmic role in reticulum stress and apoptosis,which is closely associated with the regula-tion of the PERK-Nrf2/ATF4-CHOP pathway.Conclusion Scu can effectively alleviate AAPH-induced injury to HAEC by regulating PERK-Nrf2/ATF4-CHOP pathways to inhibit endoplasmic reticulum stress and cell apoptosis.

Type 2 diabetes mellitus(T2DM)is a chronic metabolic disorder characterized primarily by pancreatic β-cell dysfunction and insulin resistance.Beyond impaired glucose homeostasis,the glucotox-icity resulting from dysregulated glucose metabolism can lead to complications such as diabetic nephropa-thy,diabetic coronary heart disease,and diabetic cognitive dysfunction,posing severe threats to patients' health.Therefore,enhancing glucose metabolism and improving insulin resistance are critical strategies for T2DM prevention and treatment.Currently,the primary therapeutic approaches for T2DM include oral hypoglycemic agents and exogenous insulin injections.However,these methods face challenges such as unclear diagnostic markers,suboptimal therapeutic efficacy,and drug-related side effects.Consequently,the development of more precise and effective treatment strategies remains a top priority in clinical re-search.MicroRNA(miRNA)is a short non-coding RNA that regulates gene transcription.It has been reported to exhibit significant alterations even before the onset of overt pathological features like hypergly-cemia,suggesting their potential as early diagnostic biomarkers for T2DM to address current issues like delayed diagnosis and disease progression due to missed optimal treatment windows.Moreover,miRNAs can modulate the expression of pathways involved in glucose metabolism,thereby improving insulin resist-ance and maintaining glucose homeostasis while avoiding side effects associated with conventional thera-pies,such as reduced bone density and localized fat accumulation.This positions miRNAs as a promising avenue for precise and efficient T2DM management.This article systematically elucidates the molecular mechanisms by which miRNAs regulate glucose metabolism-related pathways,providing a theoretical foundation and reference for future clinical research.

Objective To develop and validate a fast Monte Carlo(MC)-based patient-specific quality assurance(PSQA)tool using the treatment log files that is suitable to be used in the online adaptive radiotherapy for pencil beam scanning proton therapy(PBSPT-ART).Methods The proposed tool first used the delivery log file of a PBSPT plan to reversely reconstruct the PBSPT(rPBSPT)plan,and then used an in-house developed graphic processing unit(GPU)-accelerated virtual particle MC(VPMC)dose engine to calculate the dose distribution of the rPBSPT plan.The rPBSPT dose calculated by VPMC was then compared to the rPBSPT dose calculated by another independent MC dose engine(MCsquare),using 3D gamma analysis to verify the accuracy of VPMC calculation.As a demonstration of the feasibility of developed log file-based PSQA,the VPMC calculated dose of the rPBSPT plan was compared to the pre-delivery second check dose of the corresponding PBSPT plan calculated by MCsquare,using 3D gamma analysis.3D gamma analysis employes a criterion of 2 mm/2%/10%.Twenty patients with different disease sites were representatively selected to validate the efficiency and accuracy of the tool.Results The average calculation time of a rPBSPT plan by VPMC was(5.88±4.00)s in the accuracy verification.Compared to MCsquare,the passing rate of the 3D gamma analysis was 99.47%±0.72%.In the proposed PSQA tool demonstration,the passing rate of comparing the VPMC calculated rPBSPT dose to MCsquare calculated second check dose of the corresponding PBSPT plan was 98.91%±0.92%.Conclusion The accuracy and efficiency of the tool can meet the requirements of PSQA in the online PBSPT-ART workflow.

OBJECTIVE To investigate the epidemiological characteristics of hospital-associated infections caused by multidrug-resistant organisms(MDROs)among the burn wound patients so as to provide bases for taking tar-geted control measures.METHODS A systematic search was conducted in the worldwide database for nosocomial outbreaks,PubMed and CNKI databases so as to summarize and analyze the data regarding to outbreaks of MDROs hospital-associated infections among burn wound patients.RESULTS A total of 61 incidents of MDROs hospital-associated infections outbreaks among the burn wound patients were included,involving 2 293 patients from 21 countries and regions,50(81.97％)of which were reported for the infection sites or colonization sites in-volving burn wound,12(19.67％)involving the respiratory tract,10(16.39％)involving the bloodstream infec-tions.Methicillin-resistant Staphylococcus aureus(28 incidents,45.90％)was dominant among the pathogens causing the infections,followed by multidrug-resistant Acinetobacter baumannii(17 incidents,27.87％)and multidrug-resistant Pseudomonas aureus(9 incidents,14.75％).52 incidents(82.25％)of outbreaks were reported the contact as the major transmission mode.The suspected sources of the outbreaks included the patients(37 incidents,28.46％),health care workers(30 incidents,23.08％),ward environments(28 incidents,21.54％),medical equipments(19 incidents,30.56％),drainage systems(6 incidents,4.62％).The major pre-vention and control measures included environmental cleaning and disinfection,screening of colonization in patients and health care workers,isolation of patients with infections and hand hygiene;8 incidents were taken the measure of closing the ward.CONCLUSIONS The outbreaks of MDROs infections in the burn wound patients are mostly associated with the high frequently contact environments,medical equipments and hand hygiene of health care workers.In view of the peculiarities of the burn wound patients,it is feasible to take the targeted measures based on the summarized prevention and control combinations for MDROs so as to prevent the outbreak of hospital-asso-ciated infections.

Objective:To investigate lymph node metastasis(LNM)in the prostatic anterior fat pad(PAFP)of PCa patients after robot-assisted radical prostatectomy(RARP),and analyze the clinicopathological features and prognosis of LNM in the PAFP.Methods:We retrospectively analyzed the clinicopathological data on 1 003 cases of PCa treated by RARP in the Department of Urolo-gy of PLA General Hospital from January 2017 to December 2022.All the patients underwent routine removal of the PAFP during RARP and pathological examination,with the results of all the specimens examined and reported by pathologists.Based on the pres-ence and locations of LNM,we grouped the patients for statistical analysis,compared the clinicopathological features between different groups using the Student's t,Mann-Whitney U and Chi-square tests,and conducted survival analyses using the Kaplan-Meier and Log-rank methods and survival curves generated by Rstudio.Results:Lymph nodes were detected in 77(7.7％)of the 1 003 PAFP samples,and LNM in 11(14.3％)of the 77 cases,with a positive rate of 1.1％(11/1 003).Of the 11 positive cases,9 were found in the upgraded pathological N stage,and the other 2 complicated by pelvic LNM.The patients with postoperative pathological stage≥T3 constituted a significantly higher proportion in the PAFP LNM than in the non-PAFP LNM group(81.8％[9/11]vs 36.2％[359/992],P=0.005),and so did the cases with Gleason score ≥8(87.5％[7/8]vs 35.5％[279/786],P=0.009).No statisti-cally significant differences were observed in the clinicopathological features and biochemical recurrence-free survival between the pa-tients with PAFP LNM only and those with pelvic LNM only.Conclusion:The PAFP is a potential route to LNM,and patients with LNM in the PAFP are characterized by poor pathological features.There is no statistically significant difference in biochemical recur-rence-free survival between the patients with PAFP LNM only and those with pelvic LNM only.Routine removal of the PAFP and inde-pendent pathological examination of the specimen during RARP is of great clinical significance.

OBJECTIVE To understand the current management status of the special campaign for enhancing the pathogen detection rate before antimicrobial therapy in hospitalized patients in China,analyze identified issues dur-ing implementation,and propose improvement suggestions.METHOD A questionnaire survey was conducted to collect data from medical institutions participating in the special campaign within the region from May to Jul.2024,covering aspects such as quality control management,coordination mechanisms,data sources,indicator connotations and existing issues.RESULTS Variations were observed among medical institutions in the manage-ment of pathogen detection rates,primarily reflected in inconsistent usage of detection rate indicators,varying im-plementation levels of quality control measures and differences in multi-departmental participation.Additionally,discrepancies in data sources,statistical methods and interpretations of indicator connotations limited the compara-bility of data.CONCLUSIONS To address these issues,improvement measures such as strengthening informatiza-tion construction,standardizing indicator statistical methods and enhancing multi-departmental coordination mech-anisms should be implemented.These efforts will provide a scientific basis for the implementation of special cam-paign and robust support for the rational use of antimicrobial agents.

Objective:To evaluate the characteristic manifestations of lumbar fusion following oblique lateral interbody fusion (OLIF), determine the specific incidence of these patterns, and the identify factors associated with fusion characteristics.Methods:This retrospective study analyzed 209 patients who underwent OLIF surgery at Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, between July 2017 and September 2023. The cohort had a mean age of 64.2±9.8 years and included 125 males and 84 females. A total of 338 lumbar segments were assessed, comprising 159 segments treated with stand-alone OLIF (OLIF-SA) and 179 segments treated with OLIF combined with posterior pedicle screw fixation (OLIF-PSF). Inclusion criteria were: patients aged 18-80 years who underwent OLIF with complete radiographic records. Surgical parameters, including fixation method, number of fused segments, surgical approach, and cage dimensions (height and width), were obtained from operative records. Radiographic evaluation included preoperative osteophytes, Hounsfield unit (HU) values of endplates, and cage positioning. Fusion rate, fusion pattern (lateral vs. central), cage subsidence, and related influencing factors were assessed. Clinical outcomes were measured via the Oswestry disability index (ODI) and visual analog scale (VAS) preoperatively, immediately postoperatively, and at 1-year follow-up. Results:The overall fusion rate was 98.2% (332/338), with a non-union rate of 1.8% (6/338). The incidence of lateral fusion was 40.2% (136/338). In the OLIF-SA group, lateral and central fusion rates were 50.3% (80/159) and 49.7% (79/159), respectively, with no cases of non-union. In the OLIF-PSF group, lateral fusion occurred in 31.3% (56/179), central fusion in 65.4% (117/179), and non-union in 3.3% (6/179), with statistically significant differences between groups ( P<0.05). Preoperative osteophytes and higher endplate HU values were significantly associated with lateral fusion ( P<0.05). However, cage dimensions and cage position (anterior-posterior and lateral placement) were not significantly associated with fusion pattern ( P>0.05). Overall, 61.5% (208/338) of segments showed no cage subsidence; 24.5% (83/338) had settling, and 14.0% (47/338) had grade 1 or higher subsidence. Among lateral fusion cases, the rates of no subsidence, anchoring, grade 1, grade 2, and grade 3 subsidence were 67.6%, 21.3%, 7.4%, 3.9%, and 0.7%, respectively. In the central fusion group, these rates were 59.2%, 27.6%, 9.2%, 2.5%, and 1.3%, respectively. In the non-union group, grade 2 and 3 subsidence occurred in 50% (3/6) each, significantly higher than in the other fusion groups ( P<0.05). Post hoc analysis confirmed that grade 2 and 3 subsidence rates were significantly elevated in the non-union group compared to the lateral and central fusion groups, while other subsidence categories showed no significant differences across groups. Clinically, patients showed significant improvements in ODI and VAS scores following surgery ( P<0.05). Conclusions:Lateral fusion occurred in 40.2% of OLIF cases. The OLIF-SA technique, preoperative osteophytes, and elevated preoperative HU values were significantly associated with lateral fusion. In contrast, surgical approach, number of fused segments, cage height, width, and cage positioning did not significantly influence the occurrence of lateral fusion.

Objective To systematically evaluate the research progress of risk prediction models for periprosthetic joint infection(PJI)after total joint arthroplasty(TJA),analyze the limitations of current researches,and propose optimized suggestions.Methods Chinese and English databases such as PubMed,Embase,Web of Science,Co-chrane Library,SinoMed,Wanfang Database,VIP Database,and CNKI were retrieved systematically.The re-trieved period was from the establishment of each database to August 31,2024.Two researchers independently screened literatures and extracted data according to the CHARMS checklist,and the risk of bias in the included studies was evaluated by the PROBAST tool.Results A total of 14 studies were included in this study,involving 17 prediction models.The most common predictors included history of diabetes mellitus,obesity(body mass index[BMI]≥30 kg/m2),advanced age(≥65 years old),history of traumatic fracture,and prolonged operation time(≥2 hours).All of the included studies had high risks of bias,mainly study subject selection bias(such as single-center sample)and statistical analysis bias(such as unadjusted confounding factors).Conclusion Most of the cur-rently published risk prediction models for PJA after TJA have good predictive performance,however,there are sig-nificant limitations in the research design,especially in the insufficient control of bias risk.Future research needs to focus on improving methodological design,including adoption of prospective multi-center studies,definition of standardized predictive variables,and sufficient adjustment of confounding factors.