Atherosclerosis (AS), the primary pathological contributor to cardiovascular diseases (CVDs), has increasingly affected younger populations due to modern dietary habits and sedentary lifestyles. Current diagnostic modalities, including ultrasound, MRI, and CT, primarily identify advanced lesions and inadequately evaluate plaque vulnerability, thereby hindering early detection. Conventional treatments, which involve long-term medications associated with side effects such as hepatic injury and surgical interventions that carry risks of restenosis and hemorrhage, underscore the urgent need for non-invasive, cost-effective early diagnostic methods and targeted therapies. Gut microbiota metabolites are pivotal in AS pathogenesis, with trimethylamine N-oxide (TMAO) and short-chain fatty acids (SCFAs) serving as functionally opposing biomarkers. TMAO is produced when gut bacteria, specifically Firmicutes and Proteobacteria, metabolize dietary choline and carnitine into trimethylamine (TMA), which the liver subsequently converts to TMAO via flavin-containing monooxygenase 3 (FMO3); TMAO is then excreted in urine. Variability in TMAO levels is influenced by marine food consumption and FMO3 modulation, which can be affected by genetics, age, and diet. Mechanistically, TMAO exacerbates AS by disrupting cholesterol metabolism, inducing endothelial dysfunction through the elevation of reactive oxygen species (ROS) and pro-inflammatory cytokines such as IL-6, and reducing nitric oxide levels. Additionally, TMAO activates NF-κB and NLRP3 pathways while enhancing platelet reactivity. Clinically, elevated TMAO levels correlate with early AS and serve as predictors of mortality in patients with stable coronary artery disease (CAD) and acute coronary syndrome (ACS), as well as major adverse cardiovascular events (MACE) in stroke patients. Conversely, SCFAs—namely acetate, propionate, and butyrate—are produced by gut bacteria such as Akkermansia muciniphila and Faecalibacterium prausnitzii through the fermentation of dietary fiber. These metabolites exert anti-AS effects: acetate aids in maintaining metabolic homeostasis; propionate protects endothelial function and reduces plaque area; and butyrate fortifies intestinal barriers while suppressing inflammation. Furthermore, SCFAs cross-regulate bile acid metabolism, thereby influencing TMAO levels, and antagonize the pro-inflammatory and lipid-disrupting effects of TMAO. The use of TMAO and SCFAs as standalone biomarkers is constrained by limitations. TMAO lacks specificity, while SCFA levels fluctuate based on gut microbiota and dietary intake. Traditional AS risk assessment tools, which include clinical indicators, imaging techniques, and single biomarkers such as CRP, LDL-C, and ASCVD scores, overlook gut metabolism and demonstrate inadequate performance in younger populations. This review advocates for an “antagonistic-complementary” combined strategy: utilizing acetate and TMAO for early AS, propionate and TMAO for progressive AS, and butyrate and TMAO for advanced AS, addressing endothelial dysfunction, lipid deposition, and plaque stability/thrombosis risk, respectively. For clinical application, standardization of detection methods is crucial; liquid chromatography-mass spectrometry (LC-MS) is the gold standard, necessitating a unified sample pretreatment protocol, such as extraction with 1% formic acid in methanol. Additionally, dried blood spots (DBS) facilitate non-invasive testing, provided that dietary controls are implemented prior to detection, including a 12-hour fast and avoidance of high-choline and high-fiber foods. Existing challenges encompass the absence of standardized systems, limited large-scale validation, and ambiguous interactions with conditions such as hypertension. The authors’ team has previously established connections between gut metabolites and AS, including the reduction of TMAO as a preventive measure for AS, thereby reinforcing this proposed strategy. Future research should prioritize standardization, the development of machine learning-optimized models, validation of interventions, and the exploration of multi-omics-based “gut microbiota-metabolite-vascular” networks. In conclusion, the combined detection of TMAO and SCFAs offers a novel framework for AS risk assessment, facilitating early diagnosis and targeted interventions while enhancing the integration of gut metabolism into cardiovascular disease management.

ObjectiveMultiple sclerosis (MS) is a chronic inflammatory demyelinating disease of the central nervous system (CNS); however, its underlying neurological pathogenic mechanisms remain incompletely understood. Endogenous formaldehyde (FA), a metabolic byproduct of methylation-demethylation cycles, has recently been implicated in neurotoxicity, oxidative damage, and cognitive impairment. This study aimed to investigate whether excessive FA contributes to myelin sheath demyelination in mice and to evaluate the protective effects and mechanisms of two FA-elimination strategies: sodium bisulfite (NaHSO₃), a classical FA scavenger, and polyethylene glycol-modified astaxanthin nanoparticles (PEG-ATX@NPs), a brain-targeted nano-antioxidant formulation. MethodsA chronic demyelination model was established by feeding female C57BL/6J mice a diet containing 0.2% cuprizone (CPZ) for four weeks, followed by a two-week intervention period. Eighty mice were randomly assigned to four groups: NS (normal saline), CPZ+NS, CPZ+NaHSO₃, and CPZ+PEG-ATX@NPs. Behavioral tests, including open-field, Y-maze, and pole-climbing assays, were conducted to assess locomotor activity, motor coordination, and working memory. FA levels in serum, corpus callosum, and spinal cord were measured using an Na-FA fluorescent probe and quantified via in vivo and ex vivo fluorescence imaging. Neuroinflammatory responses were evaluated by measuring TNF-α, IL-1β, and IL-6 levels using ELISA, while oxidative stress was assessed by reactive oxygen species (ROS) fluorescence intensity. Demyelination was examined via Luxol fast blue staining, and microglial activation was analyzed by Iba1 immunofluorescence. Correlation analyses were performed to explore relationships among FA levels, inflammatory cytokines, ROS intensity, and behavioral parameters. ResultsCompared with the NS group, mice in the CPZ+NS group exhibited significant weight loss, impaired motor coordination and memory, and markedly reduced myelin regeneration (P<0.05). FA levels and pro-inflammatory cytokines were significantly elevated in serum, corpus callosum, and spinal cord (P<0.05). FA-associated fluorescence in brain and spinal tissues, as well as ROS intensity across all tissues examined, also increased substantially (P<0.05). CPZ treatment induced pronounced microglial activation and severe demyelination in the corpus callosum (P<0.01). Both NaHSO₃ and PEG-ATX@NPs effectively reduced FA accumulation in the brain and spinal cord, attenuated demyelination, suppressed microglial activation, decreased inflammatory cytokine levels, and improved motor and cognitive performance. These results confirm that CPZ induced severe demyelination accompanied by oxidative stress, neuroinflammation, and abnormal FA accumulation. Following intervention with either NaHSO₃ or PEG-ATX@NPs, endogenous FA levels in the CNS were substantially reduced. Both treatments alleviated demyelination and significantly decreased the number of activated microglia. Levels of TNF-α, IL-1β, and IL-6 in serum, corpus callosum, and spinal cord were downregulated. Behavioral performance improved significantly, as evidenced by enhanced locomotor activity, better coordination, and improved memory function. These findings indicate that both FA-scavenging agents mitigate CPZ-induced biochemical and behavioral abnormalities. ConclusionThis study demonstrates that excessive endogenous FA is closely associated with cognitive impairment, inflammatory dysregulation, and demyelination in a CPZ-induced chronic demyelination mouse model. Clearing abnormally elevated FA effectively reduces neuroinflammation, suppresses microglial overactivation, decreases oxidative stress, and alleviates demyelination, ultimately improving motor and cognitive outcomes in mice. These results suggest that targeting endogenous FA represents a promising therapeutic strategy for MS and other demyelinating disorders. Further investigations are warranted to explore the long-term safety, dosage optimization, and molecular pathways involved in FA-mediated neurotoxicity.

Atherosclerosis (AS), the primary pathological contributor to cardiovascular diseases (CVDs), has increasingly affected younger populations due to modern dietary habits and sedentary lifestyles. Current diagnostic modalities, including ultrasound, MRI, and CT, primarily identify advanced lesions and inadequately evaluate plaque vulnerability, thereby hindering early detection. Conventional treatments, which involve long-term medications associated with side effects such as hepatic injury and surgical interventions that carry risks of restenosis and hemorrhage, underscore the urgent need for non-invasive, cost-effective early diagnostic methods and targeted therapies. Gut microbiota metabolites are pivotal in AS pathogenesis, with trimethylamine N-oxide (TMAO) and short-chain fatty acids (SCFAs) serving as functionally opposing biomarkers. TMAO is produced when gut bacteria, specifically Firmicutes and Proteobacteria, metabolize dietary choline and carnitine into trimethylamine (TMA), which the liver subsequently converts to TMAO via flavin-containing monooxygenase 3 (FMO3); TMAO is then excreted in urine. Variability in TMAO levels is influenced by marine food consumption and FMO3 modulation, which can be affected by genetics, age, and diet. Mechanistically, TMAO exacerbates AS by disrupting cholesterol metabolism, inducing endothelial dysfunction through the elevation of reactive oxygen species (ROS) and pro-inflammatory cytokines such as IL-6, and reducing nitric oxide levels. Additionally, TMAO activates NF-κB and NLRP3 pathways while enhancing platelet reactivity. Clinically, elevated TMAO levels correlate with early AS and serve as predictors of mortality in patients with stable coronary artery disease (CAD) and acute coronary syndrome (ACS), as well as major adverse cardiovascular events (MACE) in stroke patients. Conversely, SCFAs—namely acetate, propionate, and butyrate—are produced by gut bacteria such as Akkermansia muciniphila and Faecalibacterium prausnitzii through the fermentation of dietary fiber. These metabolites exert anti-AS effects: acetate aids in maintaining metabolic homeostasis; propionate protects endothelial function and reduces plaque area; and butyrate fortifies intestinal barriers while suppressing inflammation. Furthermore, SCFAs cross-regulate bile acid metabolism, thereby influencing TMAO levels, and antagonize the pro-inflammatory and lipid-disrupting effects of TMAO. The use of TMAO and SCFAs as standalone biomarkers is constrained by limitations. TMAO lacks specificity, while SCFA levels fluctuate based on gut microbiota and dietary intake. Traditional AS risk assessment tools, which include clinical indicators, imaging techniques, and single biomarkers such as CRP, LDL-C, and ASCVD scores, overlook gut metabolism and demonstrate inadequate performance in younger populations. This review advocates for an “antagonistic-complementary” combined strategy: utilizing acetate and TMAO for early AS, propionate and TMAO for progressive AS, and butyrate and TMAO for advanced AS, addressing endothelial dysfunction, lipid deposition, and plaque stability/thrombosis risk, respectively. For clinical application, standardization of detection methods is crucial; liquid chromatography-mass spectrometry (LC-MS) is the gold standard, necessitating a unified sample pretreatment protocol, such as extraction with 1% formic acid in methanol. Additionally, dried blood spots (DBS) facilitate non-invasive testing, provided that dietary controls are implemented prior to detection, including a 12-hour fast and avoidance of high-choline and high-fiber foods. Existing challenges encompass the absence of standardized systems, limited large-scale validation, and ambiguous interactions with conditions such as hypertension. The authors’ team has previously established connections between gut metabolites and AS, including the reduction of TMAO as a preventive measure for AS, thereby reinforcing this proposed strategy. Future research should prioritize standardization, the development of machine learning-optimized models, validation of interventions, and the exploration of multi-omics-based “gut microbiota-metabolite-vascular” networks. In conclusion, the combined detection of TMAO and SCFAs offers a novel framework for AS risk assessment, facilitating early diagnosis and targeted interventions while enhancing the integration of gut metabolism into cardiovascular disease management.

ObjectiveMultiple sclerosis (MS) is a chronic inflammatory demyelinating disease of the central nervous system (CNS); however, its underlying neurological pathogenic mechanisms remain incompletely understood. Endogenous formaldehyde (FA), a metabolic byproduct of methylation-demethylation cycles, has recently been implicated in neurotoxicity, oxidative damage, and cognitive impairment. This study aimed to investigate whether excessive FA contributes to myelin sheath demyelination in mice and to evaluate the protective effects and mechanisms of two FA-elimination strategies: sodium bisulfite (NaHSO₃), a classical FA scavenger, and polyethylene glycol-modified astaxanthin nanoparticles (PEG-ATX@NPs), a brain-targeted nano-antioxidant formulation. MethodsA chronic demyelination model was established by feeding female C57BL/6J mice a diet containing 0.2% cuprizone (CPZ) for four weeks, followed by a two-week intervention period. Eighty mice were randomly assigned to four groups: NS (normal saline), CPZ+NS, CPZ+NaHSO₃, and CPZ+PEG-ATX@NPs. Behavioral tests, including open-field, Y-maze, and pole-climbing assays, were conducted to assess locomotor activity, motor coordination, and working memory. FA levels in serum, corpus callosum, and spinal cord were measured using an Na-FA fluorescent probe and quantified via in vivo and ex vivo fluorescence imaging. Neuroinflammatory responses were evaluated by measuring TNF-α, IL-1β, and IL-6 levels using ELISA, while oxidative stress was assessed by reactive oxygen species (ROS) fluorescence intensity. Demyelination was examined via Luxol fast blue staining, and microglial activation was analyzed by Iba1 immunofluorescence. Correlation analyses were performed to explore relationships among FA levels, inflammatory cytokines, ROS intensity, and behavioral parameters. ResultsCompared with the NS group, mice in the CPZ+NS group exhibited significant weight loss, impaired motor coordination and memory, and markedly reduced myelin regeneration (P<0.05). FA levels and pro-inflammatory cytokines were significantly elevated in serum, corpus callosum, and spinal cord (P<0.05). FA-associated fluorescence in brain and spinal tissues, as well as ROS intensity across all tissues examined, also increased substantially (P<0.05). CPZ treatment induced pronounced microglial activation and severe demyelination in the corpus callosum (P<0.01). Both NaHSO₃ and PEG-ATX@NPs effectively reduced FA accumulation in the brain and spinal cord, attenuated demyelination, suppressed microglial activation, decreased inflammatory cytokine levels, and improved motor and cognitive performance. These results confirm that CPZ induced severe demyelination accompanied by oxidative stress, neuroinflammation, and abnormal FA accumulation. Following intervention with either NaHSO₃ or PEG-ATX@NPs, endogenous FA levels in the CNS were substantially reduced. Both treatments alleviated demyelination and significantly decreased the number of activated microglia. Levels of TNF-α, IL-1β, and IL-6 in serum, corpus callosum, and spinal cord were downregulated. Behavioral performance improved significantly, as evidenced by enhanced locomotor activity, better coordination, and improved memory function. These findings indicate that both FA-scavenging agents mitigate CPZ-induced biochemical and behavioral abnormalities. ConclusionThis study demonstrates that excessive endogenous FA is closely associated with cognitive impairment, inflammatory dysregulation, and demyelination in a CPZ-induced chronic demyelination mouse model. Clearing abnormally elevated FA effectively reduces neuroinflammation, suppresses microglial overactivation, decreases oxidative stress, and alleviates demyelination, ultimately improving motor and cognitive outcomes in mice. These results suggest that targeting endogenous FA represents a promising therapeutic strategy for MS and other demyelinating disorders. Further investigations are warranted to explore the long-term safety, dosage optimization, and molecular pathways involved in FA-mediated neurotoxicity.

ObjectiveChronic atrophic gastritis (CAG) is usually diagnosed by gastroscopy and histopathological biopsy. These procedures remain the reference standard, but their invasive nature and resource requirements may limit their use in large-scale population screening and repeated follow-up. A convenient and reproducible method for noninvasive auxiliary screening may help identify individuals who require further endoscopic assessment. Fingertip photoplethysmography (PPG) provides a noninvasive recording of peripheral pulse waves and allows harmonic features to be extracted from the signal. In this study, the so-called meridian-related variables were defined as PPG-derived harmonic parameters labelled according to meridian nomenclature, rather than as direct measurements of meridian physiology. This study aimed to compare these harmonic parameters between patients with CAG and non-CAG controls, identify parameters that remained different after age adjustment, and develop a multivariable model for noninvasive auxiliary screening and pre-endoscopic risk stratification of CAG. MethodsA total of 343 participants were included, comprising 171 patients with CAG and 172 non-CAG controls. CAG diagnosis was established using gastroscopy and histopathology as the reference standard. Fingertip PPG signals were collected using a PPG-based pulse acquisition device. Eight PPG-derived harmonic parameters labelled according to meridian nomenclature were extracted for analysis. Between-group differences were first assessed using nonparametric tests. Age-adjusted analyses were then performed to reduce potential confounding by age. The false discovery rate (FDR) method was applied for multiple-comparison correction. A multivariable logistic regression model integrating age and multiple harmonic parameters was constructed. Model performance was evaluated using receiver operating characteristic (ROC) analysis and the area under the curve (AUC). Internal validation performance was assessed using stratified five-fold cross-validation and bootstrap optimism correction. Threshold performance was examined using both a high-specificity strategy and a Youden index-based cutoff. Decision curve analysis was used to evaluate the model’s net clinical benefit across a range of threshold probabilities. ResultsAll eight harmonic parameters were non-normally distributed. In the univariate analysis, the stomach-labelled harmonic parameter (ST), bladder-labelled harmonic parameter (BL), and liver-labelled harmonic parameter (LR) differed between the CAG and non-CAG groups. After age adjustment and FDR correction, only ST and BL remained statistically significant. Compared with non-CAG controls, patients with CAG showed higher ST values and lower BL values. This finding indicates an associated differential harmonic pattern that was not fully explained by age distribution. However, the discriminative ability of a single harmonic parameter was limited. The best-performing single indicator was ST, with an AUC of 0.652 (95% CI: 0.595-0.707). The multivariable model integrating age and multiple harmonic parameters achieved an AUC of 0.791 (95% CI: 0.743-0.835), representing an improvement of 0.139 over ST alone. In internal validation, stratified five-fold cross-validation yielded a mean AUC of 0.753 (95% CI: 0.715-0.781), and the bootstrap optimism-corrected AUC was 0.748. These results suggest that the model retained moderate discriminative performance after internal validation.At a specificity of at least 95%, the model achieved a sensitivity of only 40.4% (95% CI: 25.7%-49.7%). This high-specificity cutoff may be suboptimal as the preferred threshold for an initial screening setting because of the potential risk of missed CAG cases. The Youden index-based optimal cutoff was 0.419, corresponding to a sensitivity of 80.7% and a specificity of 62.8%. This threshold may better match the practical aim of noninvasive auxiliary screening, where sensitivity is usually prioritized to reduce missed cases. Decision curve analysis showed that, within a threshold probability range of 10%-55%, the model provided higher net clinical benefit than the reference strategies of recommending gastroscopy for all participants or for none. ConclusionPatients with CAG showed associated harmonic differences in fingertip PPG-derived features, mainly characterized by higher ST and lower BL values after age adjustment and FDR correction. Compared with a single harmonic parameter, the multivariable model showed better overall discrimination and retained moderate internal validation performance. These findings suggest that PPG-derived harmonic parameters labelled according to meridian nomenclature may provide auxiliary information for noninvasive auxiliary screening and front-line triage before gastroscopic confirmation in CAG. The present results support further validation rather than immediate clinical implementation. External validation in independent, multicenter, and preferably prospective screening cohorts is needed to assess the model’s generalizability, screening performance, and potential clinical utility.

ObjectiveChronic atrophic gastritis (CAG) is usually diagnosed by gastroscopy and histopathological biopsy. These procedures remain the reference standard, but their invasive nature and resource requirements may limit their use in large-scale population screening and repeated follow-up. A convenient and reproducible method for noninvasive auxiliary screening may help identify individuals who require further endoscopic assessment. Fingertip photoplethysmography (PPG) provides a noninvasive recording of peripheral pulse waves and allows harmonic features to be extracted from the signal. In this study, the so-called meridian-related variables were defined as PPG-derived harmonic parameters labelled according to meridian nomenclature, rather than as direct measurements of meridian physiology. This study aimed to compare these harmonic parameters between patients with CAG and non-CAG controls, identify parameters that remained different after age adjustment, and develop a multivariable model for noninvasive auxiliary screening and pre-endoscopic risk stratification of CAG. MethodsA total of 343 participants were included, comprising 171 patients with CAG and 172 non-CAG controls. CAG diagnosis was established using gastroscopy and histopathology as the reference standard. Fingertip PPG signals were collected using a PPG-based pulse acquisition device. Eight PPG-derived harmonic parameters labelled according to meridian nomenclature were extracted for analysis. Between-group differences were first assessed using nonparametric tests. Age-adjusted analyses were then performed to reduce potential confounding by age. The false discovery rate (FDR) method was applied for multiple-comparison correction. A multivariable logistic regression model integrating age and multiple harmonic parameters was constructed. Model performance was evaluated using receiver operating characteristic (ROC) analysis and the area under the curve (AUC). Internal validation performance was assessed using stratified five-fold cross-validation and bootstrap optimism correction. Threshold performance was examined using both a high-specificity strategy and a Youden index-based cutoff. Decision curve analysis was used to evaluate the model’s net clinical benefit across a range of threshold probabilities. ResultsAll eight harmonic parameters were non-normally distributed. In the univariate analysis, the stomach-labelled harmonic parameter (ST), bladder-labelled harmonic parameter (BL), and liver-labelled harmonic parameter (LR) differed between the CAG and non-CAG groups. After age adjustment and FDR correction, only ST and BL remained statistically significant. Compared with non-CAG controls, patients with CAG showed higher ST values and lower BL values. This finding indicates an associated differential harmonic pattern that was not fully explained by age distribution. However, the discriminative ability of a single harmonic parameter was limited. The best-performing single indicator was ST, with an AUC of 0.652 (95% CI: 0.595-0.707). The multivariable model integrating age and multiple harmonic parameters achieved an AUC of 0.791 (95% CI: 0.743-0.835), representing an improvement of 0.139 over ST alone. In internal validation, stratified five-fold cross-validation yielded a mean AUC of 0.753 (95% CI: 0.715-0.781), and the bootstrap optimism-corrected AUC was 0.748. These results suggest that the model retained moderate discriminative performance after internal validation.At a specificity of at least 95%, the model achieved a sensitivity of only 40.4% (95% CI: 25.7%-49.7%). This high-specificity cutoff may be suboptimal as the preferred threshold for an initial screening setting because of the potential risk of missed CAG cases. The Youden index-based optimal cutoff was 0.419, corresponding to a sensitivity of 80.7% and a specificity of 62.8%. This threshold may better match the practical aim of noninvasive auxiliary screening, where sensitivity is usually prioritized to reduce missed cases. Decision curve analysis showed that, within a threshold probability range of 10%-55%, the model provided higher net clinical benefit than the reference strategies of recommending gastroscopy for all participants or for none. ConclusionPatients with CAG showed associated harmonic differences in fingertip PPG-derived features, mainly characterized by higher ST and lower BL values after age adjustment and FDR correction. Compared with a single harmonic parameter, the multivariable model showed better overall discrimination and retained moderate internal validation performance. These findings suggest that PPG-derived harmonic parameters labelled according to meridian nomenclature may provide auxiliary information for noninvasive auxiliary screening and front-line triage before gastroscopic confirmation in CAG. The present results support further validation rather than immediate clinical implementation. External validation in independent, multicenter, and preferably prospective screening cohorts is needed to assess the model’s generalizability, screening performance, and potential clinical utility.

Objective To provide a scientific basis for vector prevention and control in port areas, vector monitoring has been conducted continuously to characterize the composition, density, and seasonal fluctuations of vectors across 21 port-areas of Shandong province from 2017 to 2023. Methods In accordance with the Regional Vector Monitoring Plan for Frontier Ports, vector surveillance of mosquitoes, flies, cockroaches, rodents, and other disease vectors was performed. Data were analyzed to describe vector community composition and identify seasonal fluctuations. Results A total of 601 rodents belonging to 1 family, 4 genera, and 4 species were captured. Three dominant rodent species inhabit port areas. In addition, 88 104 mosquitoes were captured and classified into 3 subfamilies, 4 genera, and 12 species; 38 959 flies were trapped and classified into 8 families,31 genera, and 51 species, including 9 dominant fly species in port-areas; and 54 672 captured cockroaches were classified into 2 families,4 genera, and 4 species, with Blattella germanica as the dominant species. Conclusions Significant vector community changes(P＜0.005)have been identified after the epidemic. To improve the pertinence and efficacy of port health quarantine work, continuous and strong attention should be paid to vector compositions and population densities at Shandong ports.

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.

Brain’s neural activities encompass both periodic rhythmic oscillations and aperiodic neural fluctuations. Rhythmic oscillations manifest as spectral peaks of neural signals, directly reflecting the synchronized activities of neural populations and closely tied to cognitive and behavioral states. In contrast, aperiodic fluctuations exhibit a power-law decaying spectral trend, revealing the multiscale dynamics of brain neural activity. In recent years, researchers have made notable progress in studying brain aperiodic dynamics. These studies demonstrate that aperiodic activity holds significant physiological relevance, correlating with various physiological states such as external stimuli, drug induction, sleep states, and aging. Aperiodic activity serves as a reflection of the brain’s sensory capacity, consciousness level, and cognitive ability. In clinical research, the aperiodic exponent has emerged as a significant potential biomarker, capable of reflecting the progression and trends of brain diseases while being intricately intertwined with the excitation-inhibition balance of neural system. The physiological mechanisms underlying aperiodic dynamics span multiple neural scales, with activities at the levels of individual neurons, neuronal ensembles, and neural networks collectively influencing the frequency, oscillatory patterns, and spatiotemporal characteristics of aperiodic signals. Aperiodic dynamics currently boasts broad application prospects. It not only provides a novel perspective for investigating brain neural dynamics but also holds immense potential as a neural marker in neuromodulation or brain-computer interface technologies. This paper summarizes methods for extracting characteristic parameters of aperiodic activity, analyzes its physiological relevance and potential as a biomarker in brain diseases, summarizes its physiological mechanisms, and based on these findings, elaborates on the research prospects of aperiodic dynamics.

Obstructive sleep apnea (OSA) is an increasingly widespread sleep-breathing disordered disease, and is an independent risk factor for many high-risk chronic diseases such as hypertension, coronary heart disease, stroke, arrhythmias and diabetes, which is potentially fatal. The key to the prevention and treatment of OSA is early diagnosis and treatment, so the assessment and diagnostic technologies of OSA have become a research hotspot. This paper reviews the research progresses of severity assessment parameters and diagnostic technologies of OSA, and discusses their future development trends. In terms of severity assessment parameters of OSA, apnea hypopnea index (AHI), as the gold standard, together with the percentage of duration of apnea hypopnea (AH%), lowest oxygen saturation (LSpO₂), heart rate variability (HRV), oxygen desaturation index (ODI) and the emerging biomarkers, constitute a multi-dimensional evaluation system. Specifically, the AHI, which measures the frequency of sleep respiratory events per hour, does not fully reflect the patients’ overall sleep quality or the extent of their daytime functional impairments. To address this limitation, the AH%, which measures the proportion of the entire sleep cycle affected by apneas and hypopneas, deepens our understanding of the impact on sleep quality. The LSpO₂ plays a critical role in highlighting the potential severe hypoxic episodes during sleep, while the HRV offers a different perspective by analyzing the fluctuations in heart rate thereby revealing the activity of the autonomic nervous system. The ODI provides a direct and objective measure of patients’ nocturnal oxygenation stability by calculating the number of desaturation events per hour, and the biomarkers offers novel insights into the diagnosis and management of OSA, and fosters the development of more precise and tailored OSA therapeutic strategies. In terms of diagnostic techniques of OSA, the standardized questionnaire and Epworth sleepiness scale (ESS) is a simple and effective method for preliminary screening of OSA, and the polysomnography (PSG) which is based on recording multiple physiological signals stands for gold standard, but it has limitations of complex operations, high costs and inconvenience. As a convenient alternative, the home sleep apnea testing (HSAT) allows patients to monitor their sleep with simplified equipment in the comfort of their own homes, and the cardiopulmonary coupling (CPC) offers a minimal version that simply analyzes the electrocardiogram (ECG) signals. As an emerging diagnostic technology of OSA, machine learning (ML) and artificial intelligence (AI) adeptly pinpoint respiratory incidents and expose delicate physiological changes, thus casting new light on the diagnostic approach to OSA. In addition, imaging examination utilizes detailed visual representations of the airway’s structure and assists in recognizing structural abnormalities that may result in obstructed airways, while sound monitoring technology records and analyzes snoring and breathing sounds to detect the condition subtly, and thus further expands our medical diagnostic toolkit. As for the future development directions, it can be predicted that interdisciplinary integrated researches, the construction of personalized diagnosis and treatment models, and the popularization of high-tech in clinical applications will become the development trends in the field of OSA evaluation and diagnosis.