Occupational cohorts are important to understanding the unique exposures of a workforce. The individuals selected by the National Aeronautics and Space Administration (NASA) to be astronauts experience occupational exposures unlike any other. To better understand the short- and long-term health effects of spaceflight, health and exposure data are collected on this cohort through clinical and other surveillance settings. This cohort is composed of the 360 astronauts who have been selected by NASA from the first selection class in 1959 to the most recent class of 2022. Selection of crewmembers is based on specific skills, education, military experience and fitness for flight. Due to the stringent and specific selection criteria, this occupational cohort encompasses a population that is more homogeneous than other groups. However, with the evolution of selection criteria along with changes to health screening and data collection processes, each selection class has a varying baseline health status. Data on a variety of health outcomes and risk factors have been collected along with occupational, physiological and exposure data, demographic and socioeconomic information and exposures that occurred prior to selection. Data have been used for both research activities, such as studies addressing Spaceflight Associated Neuro-ocular Syndrome and venous thromboembolism, and occupational surveillance activities like monitoring cardiovascular health preflight, in-flight and postflight. Characterisation of these factors helps not only with current monitoring but also informs future risk reduction decisions for exploration missions. As NASA plans missions to the Moon and Mars, the evidence base for this cohort will continue to grow through annual and mission-related data collection of current, future and retired crewmembers. Data for this cohort will continue to be collected as long as NASA continues to fly humans in space. As more data is collected, future research and surveillance activities will continue to be developed both internally and externally. One area that we aim to use this cohort is to compare with other cohorts to measure the risk of astronaut training and spaceflight, as some previous studies have done in the past.
NASA's rigorous medical selection has successfully limited in-flight medical emergencies throughout six decades of human spaceflight, but its effectiveness in identifying individuals with exceptional long-term survival remains unquantified. We estimated age-specific mortality rates based on 299 male US astronauts selected 1959-2021, using mortality data through 2022 (9602 person-years, 69 deaths). We restricted analyses to male astronauts due to insufficient follow-up time and observed mortality among female astronauts, who were first selected in 1978 and remain relatively young. We estimated astronaut hazards across the lifespan by combining rates from a Poisson regression model (for natural-cause mortality) with calculated empirical rates for external causes. These mortality rates were used to construct life tables, from which we derived life expectancy estimates and compared them to 2022 US general population values. Astronauts demonstrated substantial survival advantages at all ages, with life expectancy exceeding the general population by approximately 5 to 7 years between ages 30-70. At age 50, astronauts had a life expectancy of 36.7 years versus 29.1 years for the general population (a difference of 7.6 years). These findings quantify selection effectiveness as a countermeasure while highlighting fundamental epistemological limitations in using general population comparisons to assess spaceflight health risks.
Crewmembers on long-duration spaceflight missions are at risk of developing mild to moderate optic disc edema and ventricular volume expansion. The effect of repeat exposures on ocular and brain tissues is unknown and needs to be better understood for astronauts flying multiple missions. To evaluate whether a second exposure to spaceflight or a spaceflight analog is associated with larger changes in optic disc edema or brain ventricular volume expansion. Peripapillary total retinal thickness (TRT) extending from the Bruch membrane opening to 250 µm was quantified using optical coherence tomography imaging before and during spaceflight or the spaceflight analog head-down tilt bed rest (HDTBR). Magnetic resonance imaging was performed before and after spaceflight or HDTBR to quantify changes in brain volumetrics. Included were astronauts and participants in HDTBR from the International Space Station or German Aerospace Center :envihab facility. Study data were analyzed from November to December 2025. Two spaceflight missions of approximately 6 months or 2 HDTBR campaigns of 30 to 60 days in duration. The change in TRT (∆TRT) from preflight/pre-HDTBR to approximately 30 days before return to Earth or 30 days into HDTBR and the change in lateral ventricular volume (∆LVV) from preexposure to postexposure. Data were analyzed to determine if repeat exposure to spaceflight or HDTBR augmented the magnitude of change from the first exposure to the second. This study included a total of 7 astronauts (mean [SD] age, 43 [5] years; 5 male [71%]) and 5 participants in HDTBR (mean [SD] age, 35 [9] years; 3 male [60%]). ΔTRT was not different between spaceflight missions (mean difference, -5.6 µm; 95% CI, -15 to 3.7 µm; P = .23) or between HDTBR campaigns (mean difference, 3.1 µm; 95% CI, -3.3 to 9.5 µm; P = .33). ΔLVV was not different between spaceflight missions (mean difference, 0.1 mL; 95% CI, -0.9 to 1.1 mL; P = .78). The 3 participants in HDTBR with magnetic resonance imaging data presented with a similar ΔLVV after each campaign (0.4 vs 0.1, 1.1 vs 0.4, and 0.9 vs -0.2 mL, respectively). Findings of this case series show that a single repeat exposure to HDTBR or spaceflight did not appear to be associated with an increase in the magnitude of change in ocular or brain structures. Whether these exposures are additive in causing increased long-term functional changes remains unknown. These findings may be used by the space medicine community to guide the prediction of changes that might occur in those who undertake multiple spaceflight missions.
Bone loss occurs in astronauts during prolonged spaceflight, thus indicating the sensitivity of skeletal homeostasis to altered gravitational environments. Previous studies have shown that microgravity affects osteoclast differentiation and bone resorption, which suggests that osteoclasts possess mechanisms to sense and respond to gravity-generated mechanical forces. For testing of the related mechanisms, hypergravity can be experimentally reproduced with use of a centrifuge. In the present study, osteoclasts derived from mouse bone marrow were subjected to hypergravity under three conditions: 30G exposure using a non-CO2 centrifuge system, and short- or long-term exposure to 3G or 5G using an incubator-compatible centrifuge system. Cytoskeletal organization and resorptive function were assessed using TRAP (tartrate-resistant acid phosphatase) staining, F-actin visualization, and dentin pit assays. In addition, phosphoproteomic analysis was performed after short-term exposure to 5G hypergravity. Hypergravity exposure for as brief as 30 minutes compromised F-actin ring integrity, reduced fluorescence intensity, and promoted nuclear repositioning toward actin rings, whereas tubulin and vinculin localization remained unchanged, and the structural alterations corresponded to attenuated resorption pit formation. Quantitative phosphoproteomic profiling revealed coordinated hypergravity-dependent changes in phosphorylation across multiple cellular modules, including cytoskeletal organization, membrane trafficking, intracellular signaling, and nuclear regulatory pathways. Together, these results indicate that osteoclasts are sensitive to gravity-generated mechanical loading, with hypergravity rapidly modifying F-actin-associated cytoskeleton properties and reprogramming phosphorylation-dependent signaling networks, ultimately attenuating bone-resorptive activity. These findings provide mechanistic insight into how osteoclasts respond to altered gravitational loading conditions and have implications for skeletal adaptation during spaceflight and under altered mechanical loading conditions on Earth.
Long-duration human spaceflight will require medical systems capable of managing illness and injury without rapid evacuation or real-time assistance from Earth. Microgravity physiology, engineering limits, and communication delays reduce the feasibility of conventional surgery and favor imaging-based, minimally invasive approaches. Expeditionary interventional radiology can be defined as a practice model emphasizing image-guided, minimally invasive procedures delivered with compact equipment by small, cross-trained teams in resource-constrained environments. Research shows that astronauts and other non-specialists can obtain diagnostic-quality ultrasound images in microgravity, and analog studies demonstrate that individuals with little experience can learn key ultrasound-guided tasks after focused instruction. These findings support the feasibility of image-guided drainage, decompression, and vascular access as candidate strategies for managing acute conditions encountered during exploration missions. Remaining challenges include procedural ergonomics, equipment design, sterility, fluid containment, and development of autonomous guidance tools. This narrative review outlines a streamlined approach for adapting interventional radiology to spaceflight and highlights research needs for achieving procedural autonomy beyond Earth.
We investigate the critical phenomena of the asymmetric quantum Rabi model (AQRM), where parity symmetry is broken by an external bias. Through both analytical and numerical calculations, we identify second-order and first-order phase transitions, with the latter absent in the standard quantum Rabi model. We derive an analytical two-variable scaling function that describes the finite-frequency scaling behavior of the AQRM, and numerical results confirm this framework. The introduction of bias leads to additional critical exponents, including a bias-related critical exponent ν_{h} and susceptibility exponent γ. Moreover, we demonstrate that critical scaling persists even below the conventional critical coupling, indicating the emergence of field-induced quantum criticality. These findings establish a robust theoretical framework for understanding universal quantum criticality in light-matter systems.
Hemodialysis (HD) is the predominant treatment for end-stage renal disease (ESRD). Despite the efficacy of HD, the neurobiological underpinnings underlying high-risk complications remain unclear. In this study, using unsupervised fusion of functional and structural MRI, we identified a longitudinally altered default mode network (DMN)-insula pattern in ESRD receiving HD over 1-year follow-up (n = 39). This pattern was associated with cognition, and its related genes were enriched in biological processes involving DNA damage and repair, energy metabolism, and cellular activation. The baseline DMN-insula pattern demonstrated potential predictive value for follow-up cognition in ESRD. More importantly, these brain-cognition associations were validated in independent high-risk complications cohorts, including major depressive disorder (n = 60), mild cognitive impairment (n = 291), and Alzheimer's disease (n = 77) by extracting the corresponding brain features and assessing their correlations with cognition. Collectively, this study may help researchers better understand the underlying mechanisms of ESRD receiving HD from a multimodal neuroimaging and molecular perspective.
Spaceflight-Associated Neuro-ocular Syndrome (SANS) has emerged as a critical neuro-ophthalmic risk for human space exploration, particularly as mission duration increases and access to space expands. Current spaceflight ocular surveillance and research protocols have prioritized structural imaging and selected neuroimaging/physiological assessments. However, accumulating evidence suggests that SANS is not confined to the posterior pole as a purely structural optic nerve head phenomenon but may also involve vascular and hemodynamic alterations. At the same time, structural changes at the optic nerve head may not fully capture the functional integrity of the afferent visual pathway. We therefore propose to define a more targeted extension of current SANS surveillance protocols incorporating ultra-widefield swept-source optical coherence tomography angiography (UWF-SS-OCTA), visual evoked potentials (VEPs) and pattern electroretinogram (ERG) into standardized pre-flight, in-flight (when feasible), and post-flight assessments. Beyond its relevance to astronaut health, this topic may also be of translational interest to the broader scientific and clinical community.
Waterflooding is a widely used secondary recovery method that can substantially increase oil recovery. However, after decades of water injection, many reservoirs enter a high-water-cut stage in which large volumes of oil remain trapped, the oil fraction in the produced fluids becomes small, and overall waterflood performance deteriorates. Mobilizing this trapped oil requires a clear understanding of the types of residual oil, their spatial distribution, and the associated pore-scale formation mechanisms. Motivated by this requirement, in this work, we perform pore-scale simulations of drainage in a digital rock using the Navier-Stokes equations coupled with a volume-of-fluid method. The model and numerical implementation are validated by comparing simulated results with microchip experiments for an imbibition process. Based on the relative permeabilities obtained from the pore-scale simulations, we further derive an idealized Darcy-scale waterflood performance curve. To characterize residual oil morphology, we introduce two complementary metrics, the shape factor and Euler characteristic, and use them to classify residual oil into four categories: multipore ganglia, localized ganglia, singlets, and wetting films, and explain the corresponding formation mechanisms. Tracking the evolution of the volume fraction of each category shows that localized ganglia and singlets are the dominant trapped forms in the high-water-cut stage, together accounting for more than 80% of the residual oil. Finally, we investigate how surfactant and higher-viscosity fluid flooding mobilize residual oil in the post-waterflood regime. The additional recovery from surfactant and higher-viscosity fluid flooding is primarily associated with snap-off of multipore ganglia, driven either by reduced capillary resistance or increased viscous forcing. The volume fraction of the singlet increases markedly for both surfactant and higher-viscosity fluid flooding, and even becomes the dominant form of residual oil after higher-viscosity fluid flooding.
Catocene offers excellent catalytic activity in propellants but tends to migrate, easily leading to unstable combustion and safety hazards. To address this limitation, abundant ferrocene derivatives are synthesized by modifying the molecular polarity or enhancing interaction forces. However, most of these derivatives exist as solids and exhibit poor dispersibility in the propellant matrix. Therefore, we design a series of ferrocene-based room-temperature ionic liquids (F1-F4) as alternatives. Especially, F1 enhances weak interactions (such as hydrogen bonding and van der Waals forces) with ammonium perchlorate, thereby fundamentally suppressing migration. Moreover, owing to the retention of the active ferrocene unit, F1 is expected to exhibit excellent combustion catalytic performance comparable to catocene. Experiments confirmed that F1 could uniformly disperse within the compound solid propellant (CSP) and maintained no migration. Further, adding 2 wt % F1 increased the heat of explosion to 5373 J·g-1 (15.6% rise) and promoted the burning rate to 6.54 mm·s-1, approximately 4.0 times faster than the catalyst-free baseline (1.37 mm·s-1) and 1.9 times faster than a catocene-containing CSP (3.47 mm·s-1). The combustion temperature also rose from 1593.1 to 1972.3 °C. These results demonstrate the superior performance of F1, marking its potential as a promising burning rate catalyst substitute for catocene.
Long-duration spaceflight poses risks to musculoskeletal health, yet articular cartilage remains understudied. This review explores how microgravity and radiation compromise its homeostasis. Mechanical unloading suppresses chondrocyte metabolism and disrupts extracellular matrix equilibrium. Concurrently, radiation, oxidative stress, and immune activation induce DNA damage, mitochondrial dysfunction, and senescence, exacerbating matrix degradation. We assess physical, nutritional, and pharmacological countermeasures, highlighting the need for integrated strategies protecting joints during space exploration.
The structural design of bone scaffolds determines the local fluid mechanical microenvironment; however, how such cues program stem cell metabolism to drive osteogenesis remains unclear. In this study, Voronoi-based trabecular-like scaffolds with tunable porosity were engineered to modulate fluid shear stress (FSS) while preserving a consistent topology. Computational fluid dynamics analyses confirmed that architectures with lower porosity generated higher FSS, enabling controlled investigation of mechano-metabolic coupling. Under dynamic culture conditions, bone marrow mesenchymal stem cells (BMSCs) cultured on high-FSS scaffolds exhibited enhanced osteogenic differentiation in vitro and promoted bone regeneration in vivo. Integrated transcriptomic, proteomic, and metabolomic analyses identified caveolin-1 (CAV1) as a prominent FSS-responsive membrane regulator. Mechanistically, CAV1 enhanced phosphatidylinositol 3-kinase (PI3K)-AKT signaling, stabilized hypoxia-inducible factor-1α (HIF-1α), and induced a glycolytic shift that supports the energetic and biosynthetic demands of osteogenesis. Pharmacological inhibition of PI3K, HIF-1α, or glycolysis abolished FSS-driven osteogenic responses, validating a CAV1-centered mechano-metabolic axis. These findings establish a direct link between scaffold microarchitecture and metabolic regulation of osteogenesis and provide design principles for mechanically instructive bone repair materials.
Cancer incidence is influenced by a combination of extrinsic and genetic factors. We hypothesized that cancers with similar incidence patterns may suggest shared etiologies. Age-standardized incidence rates for 36 cancer types across 185 countries were obtained from GLOBOCAN 2022. Pairwise Spearman's correlation coefficients were computed, and network clustering analyses were performed using six community detection algorithms: Leiden, Surprise, Walktrap, Girvan-Newman, Infomap, and spectral clustering. A dominant cluster was consistently identified, comprising kidney, pancreatic, colorectal, and thyroid cancers, as well as hematological malignancies. The second cluster comprised lung cancer, mesothelioma, melanoma, and non-melanoma skin cancer, which were grouped with head and neck cancers in some algorithms. Kaposi's sarcoma, nasopharyngeal cancer, and salivary gland cancer were classified individually. The dominant cluster showed significantly greater enrichment of shared mutational signatures (cosine similarity, p = 0.041) and recurrent mutation overlap (Jaccard similarity, p = 0.025) than expected by chance. Additionally, eigenvector centrality positively correlated with global cancer incidence rates. Overall, this unsupervised network analysis of global cancer epidemiology identifies biologically coherent clusters that reflect potentially shared etiological mechanisms and may inform public health intervention strategies.
Human brain seamlessly integrates multisensory stimuli to synthesize complementary information for enhanced perceptions, depending on neural principles of superadditivity, inverse effectiveness, and temporal congruency. Replicating multisensory integration in artificial intelligences has remained challenging due to the inefficiency of algorithmic fusions and the absence of hardware-native mechanisms. Here, we demonstrate biomimetic audiovisual integration at the device level of Bi2O2Se ferroelectric-semiconductor field-effect transistors (FeS-FETs) through multiphysics coupling. Our FeS-FETs simultaneously accomplish the superadditive integration factor of 2800%, dynamical reweighting inputs of inverse effectiveness, and prolonged temporal congruency beyond 103 s. Furthermore, when configured into memristor-chip-based spiking neural networks, the resultant multisensory system is capable of executing the sensory synaptic plasticity, population-coded spiking, and Bayesian-optimal fusion, which promotes the excellent recognition accuracy of 98.2% for fuzzy objects, surpassing that identified from conventional fusion algorithms. By the exploration of multi-physical computing to mirror the biological multisensory hierarchy, we establish a physics-aware framework for neuromorphic multisensory intelligences, bridging physical dynamics with neurobiological principles for self-adaptive edge computing.
Considerable efforts have been devoted to accurately monitoring the depth of anesthesia to ensure patient safety during surgery. Traditional approaches typically rely on electroencephalogram (EEG)-based indices, such as the Bispectral Index (BIS), which require specialized equipment. In contrast, electrocardiogram (ECG) signals are widely available in clinical settings and can be conveniently acquired via wearable devices, while also exhibiting strong responsiveness to anesthetic agents. Inspired by biomimetic physiological regulation mechanisms, this study proposes a wearable-compatible ECG-based framework for depth-of-anesthesia detection that leverages autonomic nervous system characteristics and a knowledge graph-enhanced graph convolutional network (GCN). ECG recordings from 110 patients were preprocessed, and 20 anesthesia-related features were extracted, spanning morphological, statistical, spectral, heart rate variability (HRV), and entropy-based descriptors; feature selection methods identified 13 discriminative features. A patient-level knowledge graph was first constructed using the 88 training patients (1760 nodes), and test patient nodes were incorporated only after training was complete for inductive inference. Experimental results demonstrate that the proposed deep knowledge GCN achieves a test accuracy of 98.18% in distinguishing between awake and deep sleep anesthesia states, indicating that biomimetic, wearable-compatible ECG analysis combined with knowledge graph learning holds strong potential as a cost-effective alternative to traditional EEG-based anesthesia monitoring systems.
To evaluate whether mucosal boron heterogeneity measured using fluoro-boronophenylalanine positron emission tomography (18F-BPA PET) improves the prediction of oral mucositis in boron neutron capture therapy (BNCT), and to establish an imaging-guided framework for normal tissue complication probability (NTCP) modeling. This retrospective study analyzed 45 BNCT treatment sessions for head and neck cancer. Pre-treatment 18F-BPA PET was used to quantify PET-derived mucosal uptake and to derive an uptake-defined mucosal region using a tissue-to-blood ratio (TBR) threshold of 1.8. Four mucosal dose-calculation workflows were compared: (1) the Finnish workflow and (2) the Japanese workflow-both delineating mucosa anatomically and assuming uniform boron concentration implemented via a fixed TBR; (3) anatomical mucosa and (4) uptake-defined mucosa with heterogeneous boron distribution derived from PET-derived uptake heterogeneity. Biological (Gy-equivalent) and physical (Gy) subvolume dose metrics, including dose to the hottest 0.05 cubic centimeters (D0.05cc), were evaluated for correlation with mucositis severity, discrimination of grade ≥ 2 toxicity, and suitability for NTCP modeling. 18F-BPA PET demonstrated pronounced functional heterogeneity across the mucosa. Among all evaluated metrics, only the uptake-defined biological D0.05cc showed a significant monotonic association with toxicity and fulfilled all NTCP validity criteria, yielding the highest discriminative performance (area under the curve 0.765 ± 0.025) and physiologically plausible TD10-TD90 thresholds (4.85-12.84 Gy-equivalent). In contrast, Finnish and Japanese anatomical maximum-dose metrics, based on uniform-boron assumptions, showed no meaningful correlation with clinical outcomes. Functional PET imaging reveals clinically important mucosal heterogeneity that influences BNCT toxicity. The uptake-defined biological D0.05cc demonstrated superior prediction of oral mucositis and may provide a promising hypothesis-generating framework for developing patient-specific mucosal dose constraints, pending prospective validation.
Nanofluid-based spectral filtering offers a promising approach to enhance photovoltaic/thermal (PV/T) system performance by utilizing the full solar spectrum. However, system optimization remains challenging due to complex nonlinear relationships between nanofluid parameters and overall performance. This study develops a prediction-optimization framework integrating deep neural networks (DNN) with genetic algorithms (GA) to accurately analyze multi-parameter interactions and achieve globally optimal designs for nanofluid-based PV/T systems. High-throughput datasets for three nanofluids (Ag, Au, Al) were constructed using theoretical calculations that combined Lorentz-Mie theory, Monte Carlo simulations, and a coupled opto-electro-thermal model. Three machine learning models-DNN, random forest (RF), and decision tree (DT)-were employed to predict key PV/T performance parameters. By synergizing machine learning with GA, a closed-loop prediction-optimization process was established to efficiently identify optimal design parameters. Among the models evaluated, the DNN demonstrated superior performance, achieving prediction accuracies above 99.48% for all three key performance indicators (ηpv, ηth, and MF), significantly outperforming the RF and DT models. Furthermore, SHAP analysis was conducted to quantify the contribution of each input feature and enhance model interpretability. Coupled with the GA, the DNN-GA framework successfully identified globally optimal design parameters for each nanofluid. For instance, for Ag nanofluid, the optimal combination (r = 4.02 nm, h = 9.91 mm, fv = 9.45 × 10-5) yielded a maximum MF value of 1.3603. This work presents an innovative machine learning framework for designing nanofluid filters in PV/T systems, which reduces reliance on iterative experimentation and accelerates the development of high-performance solar energy systems, demonstrating practical value.
Coordination engineering of single-atom catalysts (SACs) is a powerful strategy to address durability and activity challenges in the acidic oxygen evolution reaction (OER). Here, we obtain two distinct Ir single-atom configurations on MoO3 support by regulating the second-shell coordination environment. Compared with the weakly interacting Ir─O─Mo structure, atomic pair sites formed through direct Ir─Mo coordination exhibit strong electronic coupling with the support, thereby enhancing atomic dispersion and structural stability. In situ experimental and theoretical studies reveal that the Ir─Mo pair sites trigger a new oxide-mediated pathway, in which dynamic hydroxyl spillover from Mo to Ir site effectively facilitates *OOH formation. This process breaks the linear scaling relationship between *OH and *OOH adsorption, lowering the energy barrier of the rate-limiting step and enabling superior OER kinetics. As a result, the IrO+Mo/MoO3 catalyst achieves outstanding stability for over 1500 h at 10 mA cm-2 in acidic electrolyte and sustains continuous operation for 300 h at 1.0 A cm-2 in the proton exchange membrane water electrolyzer. This work provides novel insights into the coordination engineering of SACs and opens a promising avenue for overcoming scaling limitations in acidic OER catalysis.
s: To investigate the hemodynamic consequences of renal artery ostium positioning following endovascular repair of juxtarenal aortic aneurysms (JAAAs) using the novel WeFlow-JAAA inner-branch system. Patient-specific computational fluid dynamics (CFD) analysis was performed on three postoperative JAAA cases treated with the WeFlow-JAAA endograft. For each case, four models systematically simulated progressive distal migration of the bilateral renal ostia, ranging from a proximal position near the celiac artery to their native anatomical locations. Renal perfusion rates, aortic flow patterns, and established wall shear stress (WSS)-derived metrics (time-averaged wall shear stress -TAWSS, oscillatory shear index - OSI, relative residence time - RRT) were quantitatively evaluated. Distal renal ostium positioning consistently yielded higher renal perfusion rates compared to proximal placements. However, this improved perfusion was associated with the development of potentially adverse hemodynamic conditions (characterized by low TAWSS and high OSI/RRT) on the perirenal aortic wall adjacent to the ostia. Conversely, proximal placements, while mitigating these adverse perirenal WSS patterns, compromised renal perfusion and generated pronounced flow disturbances in the infrarenal aorta distal to the termination of the inner-branch parallel segments. The preliminary findings suggest that renal ostium positioning after WeFlow-JAAA implantation may critically influence postoperative hemodynamics. The simulations indicate maximizing renal perfusion via distal placement may potentially expose the perirenal aortic wall to less favorable long-term WSS conditions. These results underscore the need for careful and potentially patient-specific consideration of device placement to appropriately balance perfusion and aortic wall remodeling risks in juxtarenal endovascular repair.
With the continuous growth of human social communication, the number of people suffering from voice disorders is also increasing. Due to the objective and non-invasive advantages of acoustic detection methods for pathological voice, the use of speech signal analysis for pathological voice recognition has become a research hotspot. This article first selected 101 continuous vowels /a/ from the German SVD database as the research object. Secondly, using wavelet packet technology for time-frequency analysis, four nonlinear dynamic parameters, namely approximate entropy, sample entropy, fuzzy entropy, and permutation entropy, are extracted from the sub signals as the feature parameter set for the pathological voice classifier. Finally, the machine learning algorithm XGBoost is selected as the pattern recognition method to establish a pathological voice classifier, and the classification performance is verified using five fold cross validation and ROC curve. Experimental results have shown that the accuracy of XGBoost's pathological voice classifier is 0.857, the F1 score is 0.875, and the AUC value is 0.944, all of which are higher than the classifier constructed by SVM, indicating that XGBoost has better performance in pathological voice recognition.