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.

ObjectiveThis study leverages structural data from antigen-antibody complexes of the influenza A virus neuraminidase (NA) protein to investigate the spatial recognition relationship between the antigenic epitopes and antibody paratopes. MethodsStructural data on NA protein antigen-antibody complexes were comprehensively collected from the SAbDab database, and processed to obtain the amino acid sequences and spatial distribution information on antigenic epitopes and corresponding antibody paratopes. Statistical analysis was conducted on the antibody sequences, frequency of use of genes, amino acid preferences, and the lengths of complementarity determining regions (CDR). Epitope hotspots for antibody binding were analyzed, and the spatial structural similarity of antibody paratopes was calculated and subjected to clustering, which allowed for a comprehensively exploration of the spatial recognition relationship between antigenic epitopes and antibodies. The specificity of antibodies targeting different antigenic epitope clusters was further validated through bio-layer interferometry (BLI) experiments. ResultsThe collected data revealed that the antigen-antibody complex structure data of influenza A virus NA protein in SAbDab database were mainly from H3N2, H7N9 and H1N1 subtypes. The hotspot regions of antigen epitopes were primarily located around the catalytic active site. The antibodies used for structural analysis were primarily derived from human and murine sources. Among murine antibodies, the most frequently used V-J gene combination was IGHV1-12*01/IGHJ2*01, while for human antibodies, the most common combination was IGHV1-69*01/IGHJ6*01. There were significant differences in the lengths and usage preferences of heavy chain CDR amino acids between antibodies that bind within the catalytic active site and those that bind to regions outside the catalytic active site. The results revealed that structurally similar antibodies could recognize the same epitopes, indicating a specific spatial recognition between antibody and antigen epitopes. Structural overlap in the binding regions was observed for antibodies with similar paratope structures, and the competitive binding of these antibodies to the epitope was confirmed through BLI experiments. ConclusionThe antigen epitopes of NA protein mainly ditributed around the catalytic active site and its surrounding loops. Spatial complementarity and electrostatic interactions play crucial roles in the recognition and binding of antibodies to antigenic epitopes in the catalytic region. There existed a spatial recognition relationship between antigens and antibodies that was independent of the uniqueness of antibody sequences, which means that antibodies with different sequences could potentially form similar local spatial structures and recognize the same epitopes.

Background: and Purpose Symptomatic intracranial hemorrhage (sICH) following endovascular thrombectomy (EVT) is a severe complication associated with adverse functional outcomes and increased mortality rates. Currently, a reliable predictive model for sICH risk after EVT is lacking. Methods: This study used data from patients aged ≥20 years who underwent EVT for anterior circulation stroke from the nationwide Taiwan Registry of Endovascular Thrombectomy for Acute Ischemic Stroke (TREAT-AIS). A predictive model including factors associated with an increased risk of sICH after EVT was developed to differentiate between patients with and without sICH. This model was compared existing predictive models using nationwide registry data to evaluate its relative performance. Results: Of the 2,507 identified patients, 158 developed sICH after EVT. Factors such as diastolic blood pressure, Alberta Stroke Program Early CT Score, platelet count, glucose level, collateral score, and successful reperfusion were associated with the risk of sICH after EVT. The TREAT-AIS score demonstrated acceptable predictive accuracy (area under the curve [AUC]=0.694), with higher scores being associated with an increased risk of sICH (odds ratio=2.01 per score increase, 95% confidence interval=1.64–2.45, P<0.001). The discriminatory capacity of the score was similar in patients with symptom onset beyond 6 hours (AUC=0.705). Compared to existing models, the TREAT-AIS score consistently exhibited superior predictive accuracy, although this difference was marginal. Conclusions The TREAT-AIS score outperformed existing models, and demonstrated an acceptable discriminatory capacity for distinguishing patients according to sICH risk levels. However, the differences between models were only marginal. Further research incorporating periprocedural and postprocedural factors is required to improve the predictive accuracy.

Background: and Purpose Symptomatic intracranial hemorrhage (sICH) following endovascular thrombectomy (EVT) is a severe complication associated with adverse functional outcomes and increased mortality rates. Currently, a reliable predictive model for sICH risk after EVT is lacking. Methods: This study used data from patients aged ≥20 years who underwent EVT for anterior circulation stroke from the nationwide Taiwan Registry of Endovascular Thrombectomy for Acute Ischemic Stroke (TREAT-AIS). A predictive model including factors associated with an increased risk of sICH after EVT was developed to differentiate between patients with and without sICH. This model was compared existing predictive models using nationwide registry data to evaluate its relative performance. Results: Of the 2,507 identified patients, 158 developed sICH after EVT. Factors such as diastolic blood pressure, Alberta Stroke Program Early CT Score, platelet count, glucose level, collateral score, and successful reperfusion were associated with the risk of sICH after EVT. The TREAT-AIS score demonstrated acceptable predictive accuracy (area under the curve [AUC]=0.694), with higher scores being associated with an increased risk of sICH (odds ratio=2.01 per score increase, 95% confidence interval=1.64–2.45, P<0.001). The discriminatory capacity of the score was similar in patients with symptom onset beyond 6 hours (AUC=0.705). Compared to existing models, the TREAT-AIS score consistently exhibited superior predictive accuracy, although this difference was marginal. Conclusions The TREAT-AIS score outperformed existing models, and demonstrated an acceptable discriminatory capacity for distinguishing patients according to sICH risk levels. However, the differences between models were only marginal. Further research incorporating periprocedural and postprocedural factors is required to improve the predictive accuracy.

Background: and Purpose Symptomatic intracranial hemorrhage (sICH) following endovascular thrombectomy (EVT) is a severe complication associated with adverse functional outcomes and increased mortality rates. Currently, a reliable predictive model for sICH risk after EVT is lacking. Methods: This study used data from patients aged ≥20 years who underwent EVT for anterior circulation stroke from the nationwide Taiwan Registry of Endovascular Thrombectomy for Acute Ischemic Stroke (TREAT-AIS). A predictive model including factors associated with an increased risk of sICH after EVT was developed to differentiate between patients with and without sICH. This model was compared existing predictive models using nationwide registry data to evaluate its relative performance. Results: Of the 2,507 identified patients, 158 developed sICH after EVT. Factors such as diastolic blood pressure, Alberta Stroke Program Early CT Score, platelet count, glucose level, collateral score, and successful reperfusion were associated with the risk of sICH after EVT. The TREAT-AIS score demonstrated acceptable predictive accuracy (area under the curve [AUC]=0.694), with higher scores being associated with an increased risk of sICH (odds ratio=2.01 per score increase, 95% confidence interval=1.64–2.45, P<0.001). The discriminatory capacity of the score was similar in patients with symptom onset beyond 6 hours (AUC=0.705). Compared to existing models, the TREAT-AIS score consistently exhibited superior predictive accuracy, although this difference was marginal. Conclusions The TREAT-AIS score outperformed existing models, and demonstrated an acceptable discriminatory capacity for distinguishing patients according to sICH risk levels. However, the differences between models were only marginal. Further research incorporating periprocedural and postprocedural factors is required to improve the predictive accuracy.

Chronic kidney disease(CKD)is a chronic disease characterized by renal structural damage and dysfunction.At present,there is still a lack of effective therapeutic drugs and prevention and treatment methods for CKD in clinical practice.More and more studies have shown that necroptosis,as a new type of programmed cell death,plays a vital role in the onset and progression of CKD.Targeting key molecules in the necroptosis pathway,such as RIPK1,RIPK3 and MLKL,the development of small molecule inhibitors has become an emerging strategy for the treatment of CKD,and has shown significant potential to pro-tect the kidneys and alleviate renal fibrosis in a variety of in vitro and in vivo models.Therefore,this article summarizes the re-search progress of the mechanism of necroptosis in recent years,and focuses on the potential role of necroptosis in the pathogene-sis of CKD and the therapeutic potential of targeting this path-way,providing a new perspective and research direction for the prevention and treatment of CKD in the future.

Objective To explore the value of support vector machine(SVM)model based on gray matter volume(GMV)for identifying amyotrophic lateral sclerosis(ALS),also to analyze the relevant brain regions.Methods MR 3D T1WI data of 60 ALS patients(ALS group)and 60 healthy volunteers(control group)were retrospectively analyzed.Taken GMV of each brain region obtained by voxel-based morphometry as the input features.F-score analysis was used to select feature with the highest classification accuracy to construct SVM model.Receiver operating characteristic curve was drawn to evaluate the efficacy of SVM model for identifying ALS,and top 10％was used as the weight threshold to obtain gray matter brain regions contributed the most to this model.Results SVM model constructed based on the top 40％GMV features had the highest classification accuracy(82.50％),with sensitivity,specificity and area under the curve(AUG)of 85.05％,80.40％and 0.890,respectively.The left precentral gyrus,left anterior cingulate gyrus and paracingulate gyrus,right middle temporal gyrus,opercular part of left inferior frontal gyrus,right dorsolateral superior frontal gyrus,left temporal pole:middle temporal gyrus,right superior occipital gyrus,orbital part of right middle frontal gyrus,right calcarine fissure and surrounding cortex,right fusiform gyrus were the top 1-10 gray matter brain regions contributed to this model.Conclusion ALS had specific GMV change pattern.SVM model based on GMV could be used to effectively identify ALS,while the left precentral gyrus was the most contributive brain region to this model.

AIM To study the chemical constituents from ethyl acetate fraction of Balanophora harlandii Hook.f.and their tyrosinase inhibitory activity.METHODS Separation and purification were performed using silica gel,MCI,ODS,Sephadex LH-20 and semi-preparative HPLC,then the structures of obtained compounds were identified by physicochemical properties and spectral data.The monophenolase inhibitory activity was determined by the tyrosinase-catalyzed oxidation of L-tyrosine.RESULTS Twenty-four compounds were isolated and identified as sesamin(1),methyl caffeate(2),quercetin(3),5,7-dihydroxychromanone(4),methyl 3,4-dihydroxybenzoate(5),esculetin(6),kaempferol(7),naringenin(8),pyrogallic acid(9),pinosylvin(10),methyl propionate(11),caffeic acid(12),saccharinol(13),ferulic acid(14),trans-p-hydroxycinnamic acid(15),cinnamic acid(16),vanillic acid(17),vanillin(18),4-hydroxyacetophenone(19),4-hydroxybenzaldehyde(20),apigenin(21),(-)-isolariciresinol(22),(-)-secoisolariciresinol(23)and meso-2,3-di(3′,4′-methylenedioxybenzyl)butane-1,4-diol(24).The IC50 values of compounds 3,5,7,8,19,and 20 ranged from(0.246 5±0.028 3)to(1.278 2±0.021 3)mmol/L.CONCLUSION Compounds 1-9、11、15、17-21、24 are isolated from this plant for the first time,and 1,6,9,17-19,24 are first isolated from genus Balanophora.Compounds 3、5、7、8、19 and 20 have tyrosinase inhibitory activity.

With the rapid development of generative artificial intelligence(GAI)technologies,their widespread application in academic research and writing is continuously expanding the boundaries of sci-entific inquiry.However,this trend has also raised a series of ethical and regulatory challenges,inclu-ding issues related to authorship,content authenticity,citation accuracy,and accountability.In light of the growing involvement of AI in generating academic content,establishing an open,controllable,and trustworthy ethical governance framework has become a key task for safeguarding research integrity and maintaining trust within the academic community.This expert consensus outlines ethical requirements across key stages of AI-assisted academic writing-including topic selection,data management,citation practices,and authorship attribution.It aims to clarify the boundaries and ethical obligations surrounding AI use in academic writing,ensuring that technological tools enhance efficiency without compromising in-tegrity.The goal is to provide guidance and institutional support for building a responsible and sustainable research ecosystem.

Aim To examine the pathogenesis of disul-fide death gene in vascular dementia(VD)by bioin-formatics analysis of disulfide death differentially ex-pressed genes(DEGs)combined with experimental verification.Methods The death DEGs of disulfide were screened and their correlation was analyzed.The VD patients data in the data set were analyzed by clus-tering and typing and gene set variation.The clustering risk of DEGs was tested with a nomogram model,and the optimal learning model was predicted.After the es-tablishment of VD rat model,water maze test,HE stai-ning and RT-qPCR detection were performed to verify the results of health information.Results Four DEGs including SLC7A11 were obtained,which had antago-nistic or synergistic interaction with each other.The genetic data could be divided into two subtypes with significant differences.After typing,VD disulfide DEGs were mainly concentrated in GnRH signaling pathways.The accuracy of the nomogram prediction model was high.Generalized linear was the best ma-chine learning model.Compared with the sham opera-tion group,the escape latency of rats in the model group was prolonged,the number of crossing platforms decreased,the relative mRNA expression levels of Slc3a2 and Slc7a11 decreased,and LRPPRC in-creased.Conclusions SLC7A11 and other disulfide death DEGs and its related GnRH signaling pathway may be an important part of the pathogenesis of VD di-sulfide death.SLC3A2,LRPPRC and SLC7A11 can be used as characteristic genes in the regulation of VD by disulfide death,which may affect VD progression through the regulation of disulfide death.