Search Results (2,184)

Medulloblastoma is a heterogeneous solid tumor, and its molecular characteristics are the most important prognostic factors for this neoplasm. Unfortunately, the molecular classification of MB-G3 and MB-G-4 medulloblastoma is very complex because of molecular similarity. Therefore, in this work, through unsupervised machine learning-based gene expression profiling, we identified a low molecular profile associated with four molecular groups of medulloblastoma. We performed medulloblastoma expression microarray data mining via the Partek Genomics Suite and Transcriptome Analysis Console (TAC), and we included a total of 25 fresh medulloblastoma tumors that were obtained and hybridized into HG U133 Plus 2.0 Array microarrays. To identify the molecular groups of the 25 patients, we compared them against classified patients, which were obtained from free repositories, and through data mining based on gene expression, compared the expression profiles of our patients. To do so, we performed an analysis via the least squares method via PCA. The molecular groups MB-WNT and MB-SHH were confirmed via immunohistochemistry via β-catenin, YAP1 and GAB1 antibodies in tissue fixed in formalin and embedded in paraffin, and another tissue section was placed on a Visium Spatial slide to perform spatial RNA-seq via Illumina NextSeq 2000 platform sequencers. The data obtained were analyzed with R. We identified the expression profiles associated with the four molecular groups and formed a reference set. Through unsupervised analysis via the least squares method, we assigned the molecular profiles of 25 patients with medulloblastoma, via the integration of bulk and spatial tumor molecular gene expression profiling analysis and with immunohistochemical findings, this strategy was fast and accurate. We observed correlations in three of the trials carried out and, in part, in one study, a patient who presented two tumor strains and two molecular signatures (SHH and G4), which led us to believe that this patient presented mixed phenotypic characteristics. Multigene expression profile analysis of medulloblastoma represents a significant advance in precision medicine; integrating different layers of transcriptomic information allows us to demonstrate underlying molecular changes in the four molecular groups that are essential for personalized therapy. Full article

Learning agility has become an essential capability for employees working in technology-driven environments characterized by rapid change and uncertainty. Despite increasing attention on learning agility, limited empirical research has examined how different levels of cognitive abilities contribute to its development, particularly among Generation Z employees. This study investigates the cognitive determinants of learning agility by distinguishing between basic cognitive abilities and high-level cognitive abilities and examining their roles across established and start-up companies. A total of 270 Generation Z employees in Indonesia participated in the study, consisting of 135 employees from established companies and 135 from start-up companies. Cognitive abilities were assessed using objective psychometric instruments, where basic cognitive abilities (reasoning, memory, attention, coordination, and perception) were measured using CogniFit, while high-level cognitive abilities were assessed through the Divergent Association Task (DAT) for creativity, the Watson–Glaser Critical Thinking Appraisal for critical thinking, and the FourSight framework for problem-solving. Learning agility was measured using a multidimensional behavioral scale. Data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results show that higher-order cognitive abilities play a more prominent role in shaping learning agility than basic cognitive abilities. Creativity and problem solving consistently demonstrate significant positive relationships with learning agility across organizational contexts, while reasoning, critical thinking, and perception show context-dependent effects across organizational environments. These findings suggest that learning agility is primarily driven by generative and evaluative cognitive processes rather than by basic cognitive efficiency alone. The study contributes to a deeper understanding of the cognitive architecture of learning agility and provides insights for organizations seeking to develop adaptive talent in rapidly evolving technological environments. Full article

Aim: Retinopathy of prematurity (ROP) is the leading cause of blindness in preterm infants. In this study, we evaluated the potential role of anemia and packed red blood cell (RBC) transfusions as risk factors in ROP development. Methods: A retrospective cohort study was conducted on premature infants who met the following inclusion criteria: infants with gestational age (GA) ≤ 32 weeks and very low birth weight (VLBW) who were admitted to the Neonatology-Preterm Department of Emergency Pediatric Hospital Cluj-Napoca during a two-year period (from 1 January 2023 to 31 December 2024). We investigated differences in the following perinatal characteristics between the two groups, those with ROP and those without: GA, birth weight (BW), severe respiratory distress syndrome, early-onset and late-onset sepsis, hemoglobin (Hb) levels, and RBC transfusions. We used the statistically significant variables to perform binary logistic regression. Results: A total of 124 newborns were recruited, with the following inclusion criteria: GA ≤ 32 weeks and BW ≤ 1500 g, of whom 79 received at least one RBC transfusion prior to 36 weeks corrected GA. Of them, 48 developed ROP with an incidence of 38.7%. In 20 cases, ROP required treatment. To adjust for clinical illness, a binary logistic regression model was created, including known risk factors for ROP and illness severity (GA, severe respiratory distress syndrome, and early- and late-onset sepsis) that were closely related to the risk of ROP development. For this regression model, Nagelkerke R-squared = 0.358, p < 0.001, and the AOR was 4.812 (95% CI: 1.374–16.847). Conclusions: RBC transfusions increased the risk of ROP. Full article

Vermicomposting is an eco-friendly biotechnology for organic waste valorization. As the primary product of earthworm biotransformation, vermicompost is a high-value bio-organic fertilizer abundant in diverse biologically active components. To date, most studies have focused on quality variation during the earthworm transformation process, while research on quality variations in the resulting vermicompost fertilizer during long-term storage remains scarce. To explore the shelf-life of vermicompost fertilizer and its key influencing indicators, this study investigated the changes in quality indicators in sealed-packaged vermicompost over a 180-day period using two typical vermicompost, namely cattle manure vermicompost (CM) and straw-amended cattle manure vermicompost (CMS). The temporal dynamics of physicochemical properties, nutrient contents, humification indices, enzyme activities, and microbial communities were monitored. The vermicompost quality was evaluated, and core quality drivers were identified using an integrated principal component analysis-partial least squares (PCA-PLS) approach. The results indicated that moisture content (MC), total organic carbon (TOC), and total nitrogen (TN) declined progressively, whereas available phosphorus (AP) and available potassium (AK) peaked at day 150 and day 120, respectively, and the humification rate (HR) increased by 2.6–4.0-fold. Bacterial diversity and relative abundance slightly decreased, accompanied by taxonomic differentiation, whereas fungal communities maintained stable diversity. Most enzyme activities, including urease, phosphatase, catalase, and dehydrogenase, reached their maxima at day 120. Comprehensive quality scores peaked at day 150, with a marked decline observed by day 180. The recommended shelf-life of vermicompost fertilizer is 150 days. The key quality determinants include TN, electrical conductivity (EC), pH, actinomycete abundance, TOC, TP, bacterial abundance, AP, AK, and HR. These findings provide theoretical support and references for the storage management and quality control of commercial vermicompost products in practice. Full article

Wildfires increasingly contribute to urban particulate matter (PM) exposure, particularly fine particles (PM_2.5), through atmospheric transport processes influenced by meteorological conditions and terrain complexity. This study investigated wildfire impacts on PM₁₀ and PM_2.5 concentrations in Split, Solin, and Kaštela (Croatia) using a terrain-aware wildfire transport framework combined with statistical and machine learning (ML) approaches. Daily PM observations (2016–2024) from three air quality monitoring stations were integrated with meteorological data from six stations, wildfire polygons, and a digital elevation model (DEM). A wildfire influence index accounting for fire size, transport distance, wind conditions, and terrain-modified airflow was evaluated using Ordinary Least Squares (OLSs) regression, Random Forest (RF) modelling, and SHAP (SHapley Additive exPlanations) analysis. Results showed stronger wildfire-related effects for PM_2.5 than for PM₁₀, while meteorological variables remained the dominant predictors of PM variability. RF models improved predictive performance relative to OLS, achieving $R^2$ = 0.474 for PM_2.5 and $R^2$ = 0.416 for PM₁₀. SHAP analysis identified precipitation, temperature, and lagged wildfire transport variables as important predictors. A total of 84 wildfire events were classified as effective wildfires, with most measurable impacts occurring within approximately 30–70 km of monitoring stations, indicating that wildfire impacts on urban air quality in Mediterranean coastal environments are strongly mediated by atmospheric transport and meteorological conditions. The proposed framework demonstrates the potential of explainable and geospatially informed ML for environmental monitoring and wildfire-related urban air quality risk assessment. Full article

The factor structure of the Utrecht Work Engagement Scale (UWES) has been debated, with studies alternately supporting unidimensional and three-factor solutions. This inconsistency may reflect a methodological limitation: conventional confirmatory factor analysis (CFA) cannot always separate general from dimension-specific variance, producing similar fit indices across competing models when a dominant general factor is present. We examined the dimensionality of the UWES-17 and UWES-9 in a sample of 755 Chilean university students, comparing unidimensional, three-factor, second-order, and bifactor models using weighted least squares mean and variance adjusted (WLSMV) estimation appropriate for ordinal data. Bifactor indices, explained common variance (ECV), percent of uncontaminated correlations (PUC), and hierarchical omega (ω_h), were computed to evaluate essential unidimensionality. Results indicated that a general engagement factor explained approximately 85% of common item variance in both versions (ECV ≈ 0.85; ω_h > 0.90), while specific factors for vigor, dedication, and absorption retained negligible reliable variance, particularly absorption (ω_h ≈ 0.00). Measurement invariance by sex was supported for the UWES-9 at the metric level, whereas classical UWES-17 solutions showed instability, including factor collapse and non-convergence of the second-order model. Taken together, findings suggest that the apparent multidimensionality of the UWES may be, at least partly, an artifact of conventional CFA modeling rather than a substantive property of the construct in this student sample. For applied monitoring of student well-being, the UWES-9 total score appears to be the most pragmatic and psychometrically defensible approach for assessing general academic engagement in this Chilean university sample, while institutional well-being monitoring would ideally be further supported by criterion-related, predictive, and sensitivity-to-change evidence. Full article

Background: Stroke remains a leading cause of mortality, disability, and health-system burden in Korea’s rapidly aging population. We aimed to describe national stroke incidence trends from 2011 to 2023, characterize ecological associations between stroke incidence and health-screening indicators, and generate model-based projections through 2028 to support health-system planning. Methods: This retrospective ecological longitudinal analysis used three publicly available aggregate national data sources: (1) NHIS annual aggregate statistics on crude and age-standardized stroke incidence, stroke case counts, first-onset vs. recurrent stroke, and case-fatality rates (2011–2023); (2) regional standardized health-awareness survey rates for stroke symptoms, myocardial infarction symptoms, blood pressure, and blood glucose (2017–2025); and (3) national cancer-screening outcome tallies for breast and cervical cancer (2010–2024). All analyses used pre-aggregated annual summary data; individual-level NHIS records were not used. Annual trends were modeled with ordinary least-squares linear regression (n = 13 annual observations). Pearson correlations were computed only for overlapping observation windows. Model-based projections are presented with 95% prediction intervals and are explicitly distinguished from observed NHIS values. This study is purely descriptive and ecological; no causal inference is made. Results: Crude stroke incidence increased from 199.2 to 221.1 per 100,000 (2011–2023; slope +2.32/year, R² = 0.83), whereas age-standardized incidence declined from 158.3 to 113.2 per 100,000 (slope −3.41/year, R² = 0.96), a pattern consistent with demographic aging as a contributing factor to growing absolute burden, though formal age-decomposition analysis would be required to confirm this inference. Total cases increased from 99,837 to 113,098; the 30-day case-fatality rate declined from 8.5% to 7.5%. Ecological correlations showed that blood glucose awareness was strongly negatively correlated with age-standardized incidence (r = −0.944, p = 0.001, n = 7), though these are ecological associations and must not be interpreted as individual-level causal relationships. Model-based projections estimate crude incidence near 230.7 (95%PI 219.2–242.2) and age-standardized incidence near 103.2 (95%PI 95.7–110.8) per 100,000 by 2026. Conclusions: Concurrent increase in crude burden and decline in age-standardized incidence reflects demographic aging as the primary driver of Korea’s stroke burden. Projections support integrated cardiovascular prevention, public health education, and age-sensitive service planning. All projections are short-horizon statistical extrapolations intended for policy scenario planning only and must not be interpreted as observed future NHIS outcomes. Full article

The seasonal monsoon reversal drives runoff and current variability along the East Asian coast, intensifying eutrophication from terrestrial nutrients. However, phytoplankton responses to these combined pressures remain poorly understood. This study analyzed their effects using partial least-squares path modeling (PLS-PM) and generalized additive models (GAMs), based on 2021 data from Shantou Bay in the Taiwan Strait, a region with complex currents and significant nutrient inputs. A total of 359 phytoplankton species were identified, with seasonal mean abundances ranging from 6.76 × 10⁶ to 57.36 × 10⁶ cells m⁻³. Ocean currents and riverine runoff drive the seasonal turnover of dominant species by modulating the temperature and salinity. In summer, the exceptionally high phytoplankton abundance in the southwestern Taiwan Strait is driven by nutrient-rich terrestrial inputs, upwelling-induced thermal inhibition, and thermocline stratification from upwelling and offshore warm waters. The phytoplankton abundance and distribution were strongly correlated with the seasonal current and runoff-driven water masses. The PLS-PM results confirm that phytoplankton dynamics are regulated by currents and terrestrial nutrient inputs altering the hydrological and chemical environments, highlighting temperature and salinity as dominant controlling factors in eutrophic coastal zones. Full article

Soil stoichiometric characteristics, as sensitive indicators of soil nutrient supply capacity and ecosystem stability, have emerged as a frontier research focus in biogeochemical cycling and ecological studies. However, the spatial variations of soil stoichiometric characteristics and driving factors in desert steppes remain unclear. Therefore, we investigated soil organic carbon (SOC), total nitrogen (TN), and total phosphorus (TP) contents and their ratios (C:N, C:P and N:P) in desert steppes in the Ili River basin, China. Results showed that: (1) in the Ili River basin, the SOC, TN, and TP contents were 30.27, 0.77, and 0.79 g·kg⁻¹, respectively, while the soil stoichiometry ratios of C:N, C:P, and N:P were 47.33, 35.48, and 1.13, respectively. All indicators demonstrated moderate variability, while soil C:P showed strong variability. (2) Significant seasonal variations were observed in SOC, TN, TP and stoichiometric ratios (p < 0.05), and soil stoichiometric characteristics were positively correlated with elevation. (3) According to Bayesian linear regression models and partial least squares-partial maximum likelihood (PLS-PM) models, climate was the principal driver of soil C, N, and their stoichiometric ratios, with mean annual temperature (MAT) and minimum temperature (Tmin) being the most influential determinants. These findings provide preliminary insights into the spatiotemporal variation patterns of soil chemical characteristics in desert steppe ecosystems of the Ili River basin. This study contributes to a deeper understanding of nutrient cycling processes within desert steppe ecosystems and offers a degree of scientific support. Full article

The shift to circular agrosystems necessitates using new ideas like sustainable biochar, which provides many eco-beneficial attributes like enhancing soil fertility, storing atmospheric carbon dioxide, and retaining soil moisture. However, there is still a small number of farmers worldwide (particularly those located in low-income countries) adopting biochar. Accordingly, this research is focused primarily on determining how factors affecting behavior will influence the decision of wheat producers in Marvdasht County, in Iran’s Fars Province, to use biochar as a circular technology for farming. The study will focus on addressing issues related to environmental challenges (e.g., degradation of soil and drought) through the implementation of resource-efficient, sustainable agricultural technologies. The intent of this paper was to research the behavioral characteristics associated with wheat farmers who choose to use biochar in the city of Marvdasht, Fars Region, Iran, using a new Theory of Planned Behavior (TPB). The model is theoretically enriched through the inclusion of personal norms and connectedness to the land, allowing for a more comprehensive understanding of pro-environmental decision-making. Data was collected from a total of 386 wheat farmers through the use of a structured survey. The data was analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) with the software Smart-PLS 3.0. The results reveal that attitude (β = 0.342, p < 0.001) and personal norms (β = 0.278, p < 0.001) are the strongest predictors of behavioral intention, while perceived behavioral control showed a weaker but significant effect (β = 0.178, p = 0.049). Subjective norms do not have a significant direct effect (β = 0.115, p = 0.199) but significantly influence intention indirectly through personal norms (β = 0.100, p < 0.001). Furthermore, connectedness to the land strongly affects personal norms (β = 0.420, p < 0.001) and exerts a significant indirect effect on intention (β = 0.117, p < 0.001), highlighting the importance of emotional attachment to land. The findings are significant because they demonstrated that farmers’ biochar adoption decisions are shaped not only by rational evaluations but also by moral obligations and emotional relationships with land. This study makes significant theoretical contributions by extending TPB with moral and relational constructs and empirically demonstrating their mediating roles in agricultural innovation adoption. The novelty of this study lies in integrating personal norms and connectedness to the land into the TPB framework to explain biochar adoption behavior within the context of circular agriculture in a developing country. Practically, the findings provide evidence-based insights for designing policies that integrate cognitive, ethical, and emotional drivers to promote biochar adoption and advance circular agriculture. Specifically, policymakers and extension agencies should prioritize behavioral-level strategies such as awareness campaigns, farmer training programs, and community-based initiatives that strengthen positive attitudes, environmental responsibility, and farmers’ emotional connection to land in order to enhance biochar adoption. Full article

How does artificial intelligence (AI) adoption shape administrative efficiency and e-citizen integration in local governments, and what role does professional development play in mediating these relationships? Drawing on a survey of 500 municipal employees across Spanish municipalities, this study employs partial least squares structural equation modelling (PLS-SEM), with formal non-linearity testing via Warp3 algorithms, to test a theoretically grounded model. The conceptual framework integrates Digital Transformation Theory and Public Value Theory as primary explanatory lenses, while drawing on the Technology Acceptance Model (TAM) and Total Factor Productivity (TFP) logic as complementary background perspectives that contextualise rather than directly operationalise the micro-level findings. Structural results reveal that AI adoption exerts a strong direct (and statistically linear) effect on perceived administrative efficiency (β = 1.04, p < 0.001; the standardised coefficient exceeding 1.0 and R² > 1 are a legitimate WarpPLS warp-model fit index rather than evidence of model misspecification: the Warp3 warp functions inflate the variance of predicted efficiency and break the additive identity SST = SSM + SSE, with the high AI–PD collinearity (r ≈ 0.84) as the contributing mechanism (RSCR = 1.000, SSR = 1.000); a comparative re-estimation without the moderation term yields β = 0.87 and R² = 0.76; we adopt this parsimonious specification (β ≈ 0.87, R² = 0.76) as the substantively interpretable estimate, with predictive relevance confirmed by a high Stone–Geisser Q² = 0.685, indicating that the model fits and predicts well rather than overfitting, while simultaneously stimulating professional development (β = 0.84, p < 0.001, R² = 0.70). Professional development positively predicted both efficiency (β = 0.27, p < 0.001) and e-citizen integration (β = 0.26, p < 0.01). Efficiency is the primary driver of e-citizen integration (β = 0.54, p < 0.001, R² = 0.53). The proposed moderation of AI adoption by professional development on efficiency was not supported (β = −0.01, p = 0.44), suggesting additive rather than synergistic effects. Model fit was robust (GoF = 0.701; ARS = 0.749; APC = 0.495); convergent and discriminant validity were confirmed by composite reliability, average variance extracted, Fornell–Larcker, and HTMT criteria; and common method bias diagnostics (Harman’s single-factor test, full-collinearity AFVIF, and marker-variable analysis) indicated that systematic method variance was not a material threat. These findings offer micro-empirical evidence of the mechanisms linking AI adoption to citizen service outcomes via a professional development pathway and provide actionable recommendations for Spanish and European municipalities navigating AI-driven governance reform. Full article

Background: Digital transformation is viewed as a lever of supply chain resilience, yet the intermediate pathways through which digital maturity relates to resilience remain underspecified, particularly in emerging-market contexts. Drawing on the Resource-Based View and the Dynamic Capabilities Framework, this study examines whether four dynamic capabilities (visibility, flexibility, risk management, and collaboration) mediate the relationship between digital maturity and supply chain resilience. Methods: Using a cross-sectional survey of 250 Moroccan firms and partial least squares structural equation modeling (PLS-SEM), we estimate a multi-mediator model and decompose the total association using variance accounted for (VAF). Results: The findings indicate that digital maturity is positively associated with resilience both directly (β = 0.219, p < 0.01) and indirectly through the four mediators, with the four capabilities jointly accounting for 63.7% of the total association (R² = 0.523, SRMR = 0.027). Visibility (18.9%) and flexibility (15.9%) emerge as the strongest indirect channels. Conclusions: The study contributes by simultaneously testing four dynamic capabilities as mediators within a single specification, documenting evidence from an under-represented emerging-market context, and providing empirically grounded managerial recommendations and policy implications. Because the data are cross-sectional, all reported coefficients describe statistical associations. Full article

Aim: This study aimed to assess the health literacy (HL) levels of patients visiting the emergency department of a university hospital and identify related factors. Methods: This cross-sectional study aimed to assess the health literacy levels of patients visiting the emergency department of a university hospital, and to identify related factors. The re-quired sample size was at least 384 individuals, assuming an inadequate HL level of 50%, with 95% confidence interval and 5% margin of error. Data were collected through a two-part questionnaire designed by the researchers. The first part covered the patients’ socio-demographic characteristics and details regarding their emergency department visits. Meanwhile, the second part included the widely used Chew’s short questions to assess inadequate HL. The analysis was conducted using IBM SPSS version 27.0. Descriptive sta-tistics, including frequency, percentage, and mean, were used to summarize the charac-teristics of the study group. The Chi-square test was applied for data analysis. Results: The study group included 58% (n = 250) female and 42% (n = 181) male. Their ages ranged from 18 to 64 years, with a mean (SD) of 29.6 (10.8) and a median of 26.0. In terms of HL levels, 197 individuals (45.7%) had inadequate HL. The frequency of inadequate HL was higher in individuals over the age of 40 years and those with an education level of ≤8 years (p < 0.05 for each). A total of 39.2% (n = 169) of the patients had visited the emergency department multiple times for their current complaints, whereas 243 participants (56.4%) visited the emergency department for a different reason within the past six months. Conclusions: In our study, four out of ten individuals had inadequate HL, and the frequency of repeated emergency department visits was quite high. No statistically significant association was found between emergency department usage characteristics and health literacy levels in the present sample, highlighting the need for larger longitudinal studies with adjusted analyses. Full article

Although the consequences of intimate partner violence (IPV) for female victims have been widely documented, the psychological and academic correlates of perpetration remain underexplored. This study examines whether emotional distress statistically mediates the association between IPV perpetration and academic presenteeism among male university students. A cross-sectional survey was administered to 343 students from the Universidad Mayor de San Andrés in Bolivia. Using validated instruments and Partial Least Squares Structural Equation Modeling, we assessed direct and indirect associations. Findings indicate that 50.1% of students reported perpetrating at least one form of IPV since entering university, with stalking and psychological violence being most common. Perpetrators reported higher levels of emotional distress compared to non-perpetrators and exhibited higher academic presenteeism (reduced academic functioning despite physical attendance). The structural model indicated a significant indirect statistical effect of IPV perpetration on academic presenteeism through emotional distress (β = 0.137, p < 0.001), accounting for 36.2% of the total effect. These findings suggest that universities may consider perpetrator-focused components within broader prevention and support systems, integrating behavioral accountability with screening, referral, and academic support while recognizing that intervention effectiveness was not tested in this study. Full article

Given the diversity and richness of China’s grain varieties, traditional physicochemical quality testing methods for rice, while providing accurate results, suffer from drawbacks such as time-consuming procedures, high costs, substantial reagent consumption, cumbersome sample preparation, and reliance on destructive or semi-destructive techniques. This study aims to employ near-infrared spectroscopy technology to establish rapid and non-destructive predictive models for key quality indicators of japonica rice. The research analyzed 133 samples from 71 widely cultivated japonica rice varieties across five major production regions in China, utilizing spectral data within a wavelength range of 660–1080 nm. Predictive models for moisture, protein, amylose, and fatty acid values were constructed using three algorithms—partial least squares regression (PLSR), ridge regression (RR), and extremely randomized trees (ERT)—linear regression and the extreme randomization tree (ERT)—their optimal parameters were determined using a 10-fold cross-validation optimization method. Eighty percent of the total dataset served as the training set, while the remaining 20% formed the test set, yielding a final test set comprising 26 samples. Performance comparisons revealed that the PLSR and RR models demonstrated superior predictive performance: the coefficient of determination (R_p²) exceeded 0.9 for all four indicators, with the R² value for fatty acid prediction reaching as high as 0.99; the root mean square error (RMSEP) of the PLSR and RR models ranged between 0.0534% and 0.3360%, confirming their high predictive accuracy. Although all ERT models (except the protein model) achieved R_p² values exceeding 0.9, their overall performance was slightly inferior to the first two methods. The protein ERT model demonstrated relatively low performance, with an R_p² value of 0.6984 on the test set, which may be attributed to the limited sample size and weak protein spectral response signals. Although the samples covered five major production regions and 71 japonica rice varieties, their distribution was uneven (multiple varieties were represented by only one or a few samples). This study provides an efficient rapid quality assessment method for japonica rice; however, the generalization ability of the models requires further validation in future studies employing larger and more balanced sample sizes. Full article