A new machine learning model to predict the prognosis of cardiogenic brain infarction.

Cardiogenic cerebral infarction (CCI) is a disease in which the blood supply to the blood vessels in the brain is insufficient due to atherosclerosis or stenosis of the coronary arteries in the patient's heart, which leads to neurological deficits. To predict the pathogenic factors of cardiogenic cerebral infarction, this paper proposes a machine learning based analytical prediction model. 494 patients with CCI who were hospitalized for the first time were consecutively included in the study between January 2017 and December 2021, and followed up every three months for one year after hospital discharge. Clinical, laboratory and imaging data were collected, and predictors associated with relapse and death in CCI patients at six months and one year after discharge were analyzed using univariate and multivariate logistic regression methods, meanwhile established a new machine learning model based on the enhanced moth-flame optimization (FTSAMFO) and the fuzzy K-nearest neighbor (FKNN), called BITSAMFO-FKNN, which is practiced on the dataset related to patients with CCI. Specifically, this paper proposes the spatial transformation strategy to increase the exploitation capability of moth-flame optimization (MFO) and combines it with the tree seed algorithm (TSA) to increase the search capability of MFO. In the benchmark function experiments FTSAMFO beat 5 classical algorithms and 5 recent variants. In the feature selection experiment, ten times ten-fold cross-validation trials showed that the BITSAMFO-FKNN model proved actual medical importance and efficacy, with an accuracy value of 96.61%, sensitivity value of 0.8947, MCC value of 0.9231, and F-Measure of 0.9444. The results of the trial showed that hemorrhagic conversion and lower LVDD/LVSD were independent risk factors for recurrence and death in patients with CCI. The established BITSAMFO-FKNN method is helpful for CCI prognosis and deserves further clinical validation.

Lithium Pollution and Its Associated Health Risks in the Largest Lithium Extraction Industrial Area in China.

Lithium (Li) is an important resource that drives sustainable mobility and renewable energy. Its demand is projected to continue to increase in the coming decades. However, the risk of Li pollution has also emerged as a global concern. Here, we investigated the pollution characteristics, sources, exposure levels, and associated health risks of Li in the Jinjiang River basin, the largest area for Li2CO3 production in China. Our results revealed the dominant role of Li extraction activities in the pollution of the river, with over 95% of dissolved Li in downstream river water being emitted from this source. Moreover, the Li concentration in aquatic plants (i.e., water hyacinth) and animals (i.e., fish) significantly increased from upstream to downstream areas, indicating a significant risk to local aquatic ecosystems. More importantly, our study found that local residents were suffering potential chronic noncarcinogenic health risks primarily from consuming contaminated water and vegetables. We also investigated the pollution characteristics of associated elements present in Li ores (e.g., Rb, Cs, Ni, and F-). By uncovering the remarkable impact of Li extraction activities on the Li content in ecosystems for the first time, our study emphasizes the importance of evaluating Li pollution from Li-related industrial activities, including mining, extraction, and recovery.

Ultrashort wave diathermy inhibits pulmonary inflammation in mice with acute lung injury in a HSP70 independent way: a pilot study.

BACKGROUND: Acute lung injury (ALI) is a clinical syndrome characterized by pulmonary inflammation. Ultrashort wave diathermy (USWD) has been shown to be effective at in inhibiting ALI inflammation, although the underlying mechanism remains unclear. Previous studies have demonstrated that USWD generates a therapeutic thermal environment that aligns with the temperature required for heat shock protein 70 (HSP70), an endogenous protective substance. In this study, we examined the correlation between HSP70 and USWD in alleviating lung inflammation in ALI. METHODS: Forty-eight male C57BL/6 mice were randomly divided into control, model, USWD intervention (LU) 1, 2, and 3, and USWD preintervention (UL) 1, 2, and 3 groups (n = 6 in each group). The mice were pretreated with LPS to induce ALI. The UL1, 2, and 3 groups received USWD treatment before LPS infusion, while the LU1, 2, and 3 groups received USWD treatment after LPS infusion. Lung function and structure, inflammatory factor levels and HSP70 protein expression levels were detected. RESULTS: USWD effectively improved lung structure and function, and significantly reduced IL-1ß, IL-10, TGF-ß1, and TNF-α levels in both the USWD preintervention and intervention groups. However, HSP70 expression did not significantly differ across the experimental groups although the expression of TLR4 was significantly decreased, suggesting that USWD may have anti-inflammatory effects through multiple signaling pathways or that the experimental conditions should be restricted. CONCLUSIONS: Both USWD intervention and preintervention effectively reduced the inflammatory response, alleviated lung injury symptoms, and played a protective role in LPS-pretreated ALI mice. HSP70 was potentially regulated by USWD in this process, but further studies are urgently needed to elucidate the correlation and mechanism.

Precisely Identifying the Sources of Magnetic Particles by Hierarchical Classification-Aided Isotopic Fingerprinting.

Magnetic particles (MPs), with magnetite (Fe3O4) and maghemite (γ-Fe2O3) as the most abundant species, are ubiquitously present in the natural environment. MPs are among the most applied engineered particles and can be produced incidentally by various human activities. Identification of the sources of MPs is crucial for their risk assessment and regulation, which, however, is still an unsolved problem. Here, we report a novel approach, hierarchical classification-aided stable isotopic fingerprinting, to address this problem. We found that naturally occurring, incidental, and engineered MPs have distinct Fe and O isotopic fingerprints due to significant Fe/O isotope fractionation during their generation processes, which enables the establishment of an Fe-O isotopic library covering complex sources. Furthermore, we developed a three-level machine learning model that not only can distinguish the sources of MPs with a high precision (94.3%) but also can identify the multiple species (Fe3O4 or γ-Fe2O3) and synthetic routes of engineered MPs with a precision of 81.6%. This work represents the first reliable strategy for the precise source tracing of particles with multiple species and complex sources.

NK cell transfer overcomes resistance to PD-(L)1 therapy in aged mice.

BACKGROUND: Cancer is the leading cause of death among older adults. Although the integration of immunotherapy has revolutionized the therapeutic landscape of cancer, the complex interactions between age and immunotherapy efficacy remain incompletely defined. Here, we aimed to elucidate the relationship between aging and immunotherapy resistance. METHODS: Flow cytometry was performed to evaluate the infiltration of immune cells in the tumor microenvironment (TME). In vivo T cell proliferation, cytotoxicity and migration assays were performed to evaluate the antitumor capacity of tumor antigen-specific CD8+ T cells in mice. Real-time quantitative PCR (qPCR) was used to investigate the expression of IFN-γ-associated gene and natural killer (NK)-associated chemokine. Adoptive NK cell transfer was adopted to evaluate the effects of NK cells from young mice in overcoming the immunotherapy resistance of aged mice. RESULTS: We found that elderly patients with advanced non-small cell lung cancer (aNSCLC) aged ≥ 75 years exhibited poorer progression-free survival (PFS), overall survival (OS) and a lower clinical response rate after immunotherapy. Mechanistically, we showed that the infiltration of NK cells was significantly reduced in aged mice compared to younger mice. Furthermore, the aged NK cells could also suppress the activation of tumor antigen-specific CD8+ T cells by inhibiting the recruitment and activation of CD103+ dendritic cells (DCs). Adoptive transfer of NK cells from young mice to aged mice promoted TME remodeling, and reversed immunotherapy resistance. CONCLUSION: Our findings revealed the decreased sensitivity of elderly patients to immunotherapy, as well as in aged mice. This may be attributed to the reduction of NK cells in aged mice, which inhibits CD103+ DCs recruitment and its CD86 expression and ultimately leads to immunotherapy resistance.

Use of the Chinese version of the MATRICS Consensus Cognitive Battery to assess cognitive functioning in individuals with high risk for psychosis, first-episode schizophrenia and chronic schizophrenia: a systematic review and meta-analysis.

More than one hundred studies have used the mainland Chinese version of the MATRICS Consensus Cognitive Battery (MCCB) to assess cognition in schizophrenia, but the results of these studies, the quality of the reports, and the strength of the evidence provided in the reports have not been systematically assessed. We identified 114 studies from English-language and Chinese-language databases that used the Chinese MCCB to assess cognition in combined samples of 7394 healthy controls (HC), 392 individuals with clinical high risk for psychosis (CHR-P), 4922 with first-episode schizophrenia (FES), 1549 with chronic schizophrenia (CS), and 2925 with schizophrenia of unspecified duration. The mean difference (MD) of the composite MCCB T-score (-13.72) and T-scores of each of the seven cognitive domains assessed by MCCB (-14.27 to -7.92) were significantly lower in individuals with schizophrenia than in controls. Meta-analysis identified significantly greater cognitive impairment in FES and CS than in CHR-P in six of the seven domains and significantly greater impairment in CS than FES in the reasoning and problem-solving domain (i.e., executive functioning). The only significant covariate of overall cognitive functioning in individuals with schizophrenia was a negative association with the severity of psychotic symptoms. These results confirm the construct validity of the mainland Chinese version of MCCB. However, there were significant limitations in the strength of the evidence provided about CHR-P (small pooled sample sizes) and the social cognition domain (inconsistency of results across studies), and the quality of many reports (particularly those published in Chinese) was rated 'poor' due to failure to report sample size calculations, matching procedures or methods of handling missing data. Moreover, almost all studies were cross-sectional studies limited to persons under 60 with at least nine years of education, so longitudinal studies of under-educated, older individuals with schizophrenia are needed.

Design, synthesis and neuroprotective biological evaluation of novel HDAC6 inhibitors incorporating benzothiadiazinyl systems as cap groups.

Histone deacetylase 6 (HDAC6), as the key regulatory enzyme, plays an important role in the development of the nervous system. More and more studies indicate that HDAC6 has become a promising therapeutic target for CNS diseases. Herein we designed and synthesized a series of novel HDAC6 inhibitors with benzothiadiazinyl systems as cap groups and evaluated their activity in vitro and in vivo. Among them, compound 3 exhibited superior selective inhibitory activity against HDAC6 (IC50 = 5.1 nM, about 30-fold selectivity over HDAC1). The results of docking showed that compound 3 can interact well with the key amino acid residues of HDAC6. Compound 3 showed lower cytotoxicity (20 µM to SH-SY5Y cells, inhibition rate = 25.75%) and better neuroprotective activity against L-glutamate-induced SH-SY5Y cell injury model in vitro. Meanwhile, compound 3 exhibited weak cardiotoxicity (10 µM hERG inhibition rate = 17.35%) and possess good druggability properties. Especially, compound 3 could significantly reduce cerebral infarction from 49.87% to 32.18%, and similar with butylphthalide in MCAO model, indicating potential clinical application prospects for alleviating ischemic stroke-induced brain infarction.

Circulating T cells: a promising biomarker of anti-PD-(L)1 therapy.

Anti-PD-(L)1 therapy has shown great efficacy in some patients with cancer. However, a significant proportion of patients with cancer do not respond to it. Another unmet clinical need for anti-PD-(L)1 therapy is the dynamic monitoring of treatment effects. Therefore, identifying biomarkers that can stratify potential responders before PD-(L)1 treatment and timely monitoring of the efficacy of PD-(L)1 treatment are crucial in the clinical setting. The identification of biomarkers by liquid biopsy has attracted considerable attention. Among the identified biomarkers, circulating T cells are one of the most promising because of their indispensable contribution to anti-PD-(L)1 therapy. The present review aimed to thoroughly explore the potential of circulating T cells as biomarkers of anti-PD-(L)1 therapy and its advantages and limitations.

Preserving shape details of pulse signals for video-based blood pressure estimation.

In recent years, imaging photoplethysmograph (iPPG) pulse signals have been widely used in the research of non-contact blood pressure (BP) estimation, in which BP estimation based on pulse features is the main research direction. Pulse features are directly related to the shape of pulse signals while iPPG pulse signals are easily disturbed during the extraction process. To mitigate the impact of pulse feature distortion on BP estimation, it is necessary to eliminate interference while retaining valuable shape details in the iPPG pulse signal. Contact photoplethysmograph (cPPG) pulse signals measured at rest can be considered as the undisturbed reference signal. Transforming the iPPG pulse signal to the corresponding cPPG pulse signal is a method to ensure the effectiveness of shape details. However, achieving the required shape accuracy through direct transformation from iPPG to the corresponding cPPG pulse signals is challenging. We propose a method to mitigate this challenge by replacing the reference signal with an average cardiac cycle (ACC) signal, which can approximately represent the shape information of all cardiac cycles in a short time. A neural network using multi-scale convolution and self-attention mechanisms is developed for this transformation. Our method demonstrates a significant improvement in the maximal information coefficient (MIC) between pulse features and BP values, indicating a stronger correlation. Moreover, pulse signals transformed by our method exhibit enhanced performance in BP estimation using different model types. Experiments are conducted on a real-world database with 491 subjects in the hospital, averaging 60 years of age.

Probing the stability of metal-organic frameworks by structure-responsive mass spectrometry imaging.

The widespread application of metal-organic frameworks (MOFs) is seriously hindered by their structural instability and it is still very challenging to probe the stability of MOFs during application by current techniques. Here, we report a novel structure-responsive mass spectrometry (SRMS) imaging technique to probe the stability of MOFs. We discovered that intact CuBTC (as a model of MOFs) could generate the characteristic peaks of organic ligands and carbon cluster anions in laser desorption/ionization mass spectrometry, but these peaks were significantly changed when the structure of CuBTC was dissociated, thus enabling a label-free probing of the stability. Furthermore, SRMS can be performed in imaging mode to visualize the degradation kinetics and reveal the spatial heterogeneity of the stability of CuBTC. This technique was successfully applied in different application scenarios (in water, moist air, and CO2) and also validated with different MOFs. It thus provides a versatile new tool for better design and application of environment-sensitive materials.

Predicting Arterial Stiffness From Single-Channel Photoplethysmography Signal: A Feature Interaction-Based Approach.

Arterial stiffness (AS) serves as a crucial indicator of arterial elasticity and function, typically requiring expensive equipment for detection. Given the strong correlation between AS and various photoplethysmography (PPG) features, PPG emerges as a convenient method for assessing AS. However, the limitations of independent PPG features hinder detection accuracy. This study introduces a feature selection method leveraging the interactive relationships between features to enhance the accuracy of predicting AS from a single-channel PPG signal. Initially, an adaptive signal interception method was employed to capture high-quality signal fragments from PPG sequences. 58 PPG features, deemed to have potential contributions to AS estimation, were extracted and analyzed. Subsequently, the interaction factor (IF) was introduced to redefine the interaction and redundancy between features. A feature selection algorithm (IFFS) based on the IF was then proposed, resulting in a combination of interactive features. Finally, the Xgboost model is utilized to estimate AS from the selected features set. The proposed approach is evaluated on datasets of 268 male and 124 female subjects, respectively. The results of AS estimation indicate that IFFS yields interacting features from numerous sources, rejects redundant ones, and enhances the association. The interaction features combined with the Xgboost model resulted in an MAE of 122.42 and 142.12 cm/sec, an SDE of 88.16 and 102.56 cm/sec, and a PCC of 0.88 and 0.85 for the male and female groups, respectively. The findings of this study suggest that the stated method improves the accuracy of predicting AS from single-channel PPG, which can be used as a non-invasive and cost-effective screening tool for atherosclerosis.

HRUNet: Assessing Uncertainty in Heart Rates Measured From Facial Videos.

Video-based Photoplethysmography (VPPG) offers the capability to measure heart rate (HR) from facial videos. However, the reliability of the HR values extracted through this method remains uncertain, especially when videos are affected by various disturbances. Confronted by this challenge, we introduce an innovative framework for VPPG-based HR measurements, with a focus on capturing diverse sources of uncertainty in the predicted HR values. In this context, a neural network named HRUNet is structured for HR extraction from input facial videos. Departing from the conventional training approach of learning specific weight (and bias) values, we leverage the Bayesian posterior estimation to derive weight distributions within HRUNet. These distributions allow for sampling to encode uncertainty stemming from HRUNet's limited performance. On this basis, we redefine HRUNet's output as a distribution of potential HR values, as opposed to the traditional emphasis on the single most probable HR value. The underlying goal is to discover the uncertainty arising from inherent noise in the input video. HRUNet is evaluated across 1,098 videos from seven datasets, spanning three scenarios: undisturbed, motion-disturbed, and light-disturbed. The ensuing test outcomes demonstrate that uncertainty in the HR measurements increases significantly in the scenarios marked by disturbances, compared to that in the undisturbed scenario. Moreover, HRUNet outperforms state-of-the-art methods in HR accuracy when excluding HR values with 0.4 uncertainty. This underscores that uncertainty emerges as an informative indicator of potentially erroneous HR measurements. With enhanced reliability affirmed, the VPPG technique holds the promise for applications in safety-critical domains.

Metabolite-sensing GPCRs in rheumatoid arthritis.

Persistent inflammation in damaged joints results in metabolic dysregulation of the synovial microenvironment, causing pathogenic alteration of cell activity in rheumatoid arthritis (RA). Recently, the role of metabolite and metabolite-sensing G protein-coupled receptors (GPCRs) in the RA-related inflammatory immune response (IIR) has become a focus of research attention. These GPCRs participate in the progression of RA by modulating immune cell activation, migration, and inflammatory responses. Here, we discuss recent evidence implicating metabolic dysregulation in RA pathogenesis, focusing on the connection between RA-related IIR and GPCR signals originating from the synovial joint and gut. Furthermore, we discuss future directions for targeting metabolite-sensing GPCRs for therapeutic benefit, emphasizing the importance of identifying endogenous ligands and investigating the various transduction mechanisms involved.

Mechanisms of rainfall-induced landslides and interception dynamic response: a case study of the Ni changgou landslide in Shimian, China.

Geological hazards, especially landslides and mudslides, are frequent in Caoke County, Sichuan Province, China. In September 2022, the mechanical parameters of the soil were obtained through a basic investigation of the landslide characteristics of Ni changgou. Upon that, the finite element-discrete element method was used to reconstruct the three-dimensional numerical model of the landslide on the right bank of Ni changgou, and the initiation mechanism of rainfall on landslide and the formation of debris flow impact dam process were simulated. Furthermore, the pore pressure, stability coefficient as well as displacement of the landslide body were analyzed. It turned out that with the increase of rainfall intensity, the pore water pressure value also increases, where pore water pressure rises rapidly. the slope is close to the unstable edge, Eventually, it tends to one under rainfall conditions, and due to gravity, the slide of the landslide is induced. The duration of landslide movement is about 200 s, the maximum average velocity of the landslide reaches 4.85 m/s, and the average movement distance is close to 500 m. In addition, this method is applied to the Chutougou debris flow, and the corresponding hazard analysis is added which could better show the treatment and application of debris flow in actual engineering.

GRK2 inhibits Flt-1+ macrophage infiltration and its proangiogenic properties in rheumatoid arthritis.

Rheumatoid arthritis (RA) is an autoimmune disease with a complex etiology. Monocyte-derived macrophages (MDMs) infiltration are associated with RA severity. We have reported the deletion of G-protein-coupled receptor kinase 2 (GRK2) reprograms macrophages toward an anti-inflammatory phenotype by recovering G-protein-coupled receptor signaling. However, as more GRK2-interacting proteins were discovered, the GRK2 interactome mechanisms in RA have been understudied. Thus, in the collagen-induced arthritis mouse model, we performed genetic GRK2 deletion using GRK2f/fLyz2-Cre+/- mice. Synovial inflammation and M1 polarization were improved in GRK2f/fLyz2-Cre+/- mice. Supporting experiments with RNA-seq and dual-luciferase reporter assays identified peroxisome proliferator-activated receptor γ (PPARγ) as a new GRK2-interacting protein. We further confirmed that fms-related tyrosine kinase 1 (Flt-1), which promoted macrophage migration to induce angiogenesis, was inhibited by GRK2-PPARγ signaling. Mechanistically, excess GRK2 membrane recruitment in CIA MDMs reduced the activation of PPARγ ligand-binding domain and enhanced Flt-1 transcription. Furthermore, the treatment of mice with GRK2 activity inhibitor resulted in significantly diminished CIA pathology, Flt-1+ macrophages induced-synovial inflammation, and angiogenesis. Altogether, we anticipate to facilitate the elucidation of previously unappreciated details of GRK2-specific intracellular signaling. Targeting GRK2 activity is a viable strategy to inhibit MDMs infiltration, affording a distinct way to control joint inflammation and angiogenesis of RA.

Involvement of Embryo-Derived and Monocyte-Derived Intestinal Macrophages in the Pathogenesis of Inflammatory Bowel Disease and Their Prospects as Therapeutic Targets.

Inflammatory bowel disease (IBD) is a group of intestinal inflammatory diseases characterized by chronic, recurrent, remitting, or progressive inflammation, which causes the disturbance of the homeostasis between immune cells, such as macrophages, epithelial cells, and microorganisms. Intestinal macrophages (IMs) are the largest population of macrophages in the body, and the abnormal function of IMs is an important cause of IBD. Most IMs come from the replenishment of blood monocytes, while a small part come from embryos and can self-renew. Stimulated by the intestinal inflammatory microenvironment, monocyte-derived IMs can interact with intestinal epithelial cells, intestinal fibroblasts, and intestinal flora, resulting in the increased differentiation of proinflammatory phenotypes and the decreased differentiation of anti-inflammatory phenotypes, releasing a large number of proinflammatory factors and aggravating intestinal inflammation. Based on this mechanism, inhibiting the secretion of IMs' proinflammatory factors and enhancing the differentiation of anti-inflammatory phenotypes can help alleviate intestinal inflammation and promote tissue repair. At present, the clinical medication of IBD mainly includes 5-aminosalicylic acids (5-ASAs), glucocorticoid, immunosuppressants, and TNF-α inhibitors. The general principle of treatment is to control acute attacks, alleviate the condition, reduce recurrence, and prevent complications. Most classical IBD therapies affecting IMs function in a variety of ways, such as inhibiting the inflammatory signaling pathways and inducing IM2-type macrophage differentiation. This review explores the current understanding of the involvement of IMs in the pathogenesis of IBD and their prospects as therapeutic targets.

VidBP: Detecting Blood Pressure from Facial Videos with Personalized Calibration.

Recent studies have found that blood volume pulse (BVP) in facial videos contains features highly correlated to blood pressure (BP). However, the mapping from BVP features to BP varies from person to person. To address this issue, VidBP has been proposed as a BP detector that can be calibrated based on an individual's data. VidBP is pre-trained on a large dataset to extract BP-related features from BVP. Then, BVP samples and BP labels of an individual are fed into the pre-trained VidBP to create a personal dictionary of BP-related features. When estimating the individual's BP, the current BP-related feature is compared to the features saved in the dictionary, and the BP labels of the similar features are considered as the BP estimate. The performance of VidBP was evaluated on 640 samples of 16 subjects, and it demonstrated significantly lower errors in BP estimation compared to state-of-the-art methods. The personalized calibration of VidBP is a significant advantage, enabling it to better capture the unique mapping from BVP features to BP for each individual.Clinical relevance This study reports a feasible method to estimate BP from facial videos, providing a convenient and cost-effective way for home BP monitoring.

Perspectives and new aspects of histone deacetylase inhibitors in the therapy of CNS diseases.

Many populations worldwide are suffering from central nervous system (CNS) diseases such as brain tumors, neurodegenerative diseases (Alzheimer's disease, Parkinson's disease and Huntington's disease) and stroke. There is a shortage of effective drugs for most CNS diseases. As one of the regulatory mechanisms of epigenetics, the particular role and therapeutic benefits of histone deacetylases (HDACs) in the CNS have been extensively studied. In recent years, HDACs have attracted increasing attention as potential drug targets for CNS diseases. In this review, we summarize the recent applications of representative histone deacetylases inhibitors (HDACis) in CNS diseases and discuss the challenges in developing HDACis with different structures and better blood-brain barrier (BBB) permeability, hoping to promote the development of more effective bioactive HDACis for the treatment of CNS diseases.

Extracellular matrix in synovium development, homeostasis and arthritis disease.

Extracellular matrix (ECM) is a three-dimensional network entity composed of extracellular macromolecules. ECM in synovium not only supports the structural integrity of synovium, but also plays a crucial role in regulating homeostasis and damage repair response in synovium. Obvious disorders in the composition, behavior and function of synovial ECM will lead to the occurrence and development of arthritis diseases such as rheumatoid arthritis (RA), osteoarthritis (OA) and psoriatic arthritis (PsA). Based on the importance of synovial ECM, targeted regulation of the composition and structure of ECM is considered to be an effective measure for the treatment of arthritis disease. This paper reviews the current research status of synovial ECM biology, discusses the role and mechanism of synovial ECM in physiological status and arthritis disease, and summarizes the current strategies for targeting synovial ECM to provide information for the pathogenesis, diagnosis and treatment of arthritis disease.

PFDNet: A Pulse Feature Disentanglement Network for Atrial Fibrillation Screening From Facial Videos.

Video-based Photoplethysmography (VPPG) can identify arrhythmic pulses during atrial fibrillation (AF) from facial videos, providing a convenient and cost-effective way to screen for occult AF. However, facial motions in videos always distort VPPG pulse signals and thus lead to the false detection of AF. Photoplethysmography (PPG) pulse signals offer a possible solution to this problem due to the high quality and resemblance to VPPG pulse signals. Given this, a pulse feature disentanglement network (PFDNet) is proposed to discover the common features of VPPG and PPG pulse signals for AF detection. Taking a VPPG pulse signal and a synchronous PPG pulse signal as inputs, PFDNet is pre-trained to extract the motion-robust features that the two signals share. The pre-trained feature extractor of the VPPG pulse signal is then connected to an AF classifier, forming a VPPG-driven AF detector after joint fine-tuning. PFDNet has been tested on 1440 facial videos of 240 subjects (50% AF absence and 50% AF presence). It achieves a Cohen's Kappa value of 0.875 (95% confidence interval: 0.840-0.910, P<0.001) on the video samples with typical facial motions, which is 6.8% higher than that of the state-of-the-art method. PFDNet shows significant robustness to motion interference in the video-based AF detection task, promoting the development of opportunistic screening for AF in the community.