Browse Articles

Correction: An intercomparison of models predicting growth of Antarctic krill (Euphausia superba): The importance of recognizing model specificity

Repeatability of flicker modulation sensitivity measured using the Flicker-Plus test

Purpose To evaluate the test-retest variability of the Flicker-Plus test for each of the two protocols measuring rod and cone-enhanced flicker modulation thresholds (FMT) in healthy individuals. A secondary aim was to evaluate the within-subject variability in repeated measurements. Methods Thirty healthy participants aged 19–71 years were examined. None had any history or signs of ocular disease. Monocular FMT were measured at the fovea (0°) and at an eccentricity of 5° in each quadrant, twice by the same investigator under identical conditions within a 2-week period under stimulus conditions that favoured either rods or cones to evaluate the between session repeatability. To assess the within-subject variability, binocular measurements for cone and rod-enhanced FMT were carried out on 15 different occasions over a period of 3-weeks in three of the participants. Coefficient of Repeatability (CoR) was calculated for inter-session repeatability and Bland Altman plots were created for graphical representation. Inter-class correlation coefficient (ICC) with 95% confidence intervals were estimated. Results Bland and Altman analysis shows that the mean bias is greater than zero in all 5 testing locations for both rod and cone-enhanced FMTs, suggesting that the threshold at the second visit tended to be lower than at the first, however the difference between visits was not statistically significant for any test condition (paired t-test, p < 0.05). In a sub analysis for those CoR was found to be higher in those aged <45years, compared to those aged ≥45 years. The correlation and agreement between the two measurements estimated by ICC analysis shows good (0.75–0.9) to excellent (>0.9) test-retest reliability of Flicker-Plus test for all measures. Conclusion The findings show good to excellent test-retest repeatability for the Flicker-Plus test. This was the case at all locations (foveal and peripheral), under both cone and rod-enhanced conditions. There was no evidence of significant learning effects.

Immune subtyping of lymph node metastasis-negative colorectal cancer reveals biomarkers for prognosis and immunotherapy response

Background Lymph node metastasis (LNM) is a key prognostic factor in colorectal cancer (CRC), and early lymph node status assessment is crucial for prognosis and immunotherapy decisions. However, the immune characteristics of LNM-negative CRC remain poorly understood. Methods Using machine learning algorithms, we identified and analyzed two immune subtypes (C1 and C2) in 244 LNM-negative CRC samples, with validation in 458 additional samples, and evaluated their immune characteristics and functional pathways. Results Subtype C1 exhibited high immune scores and responsiveness to immune checkpoint blockade, while subtype C2 showed a favorable prognosis and increased immune cell infiltration, indicating it represents an earlier CRC stage. Experimental validation revealed that PPFIA4 knockdown in C1 significantly suppressed CRC cell proliferation and migration. Conclusions These findings provide insights into personalized immunotherapy strategies for early-stage CRC patients and have potential clinical application value.

The random noise modulations on the nonlinear Chiral Schrödinger structures

In this paper, we consider the Chiral nonlinear Schödinger equation (CNLSE), where the multiplicative noises term varies arbitrarily over time. This equation defines several edge states of Hall effect characteristics in quantum physics applications. We apply the sine-Gordon expansion method to produce some new stochastic solutions for the CNLSE. Some solitary and dissipative solutions were obtained in the form of rational, envelope and shock structures. We demonstrate how the multiplicative noise and model parameters affects the way the solutions behave. We provide some configurations for the both deterministic and stochastic solutions to illustrate their behaviour. It is known that noise dominates envelope growing, damping, and all wave propagation. As it is achieved, the results presented here are crucial to the development of quantum physics. The proposed methodology can be developed to solve more complex problems in applied science.

ELFT: Efficient local-global fusion transformer for small object detection

Small object detection is an essential but challenging task in computer vision. Transformer-based algorithms have demonstrated remarkable performance in the domain of computer vision tasks. Nevertheless, they suffer from inadequate feature extraction for small objects. Additionally, they face difficulties in deployment on resource-constrained platforms due to their heavy computational burden. To tackle these problems, an efficient local-global fusion Transformer (ELFT) is proposed for small object detection, which is based on attention and grouping strategy. Specifically, we first design an efficient local-global fusion attention (ELGFA) mechanism to extract sufficient location features and integrate detailed information from feature maps, thereby promoting the accuracy. Besides, we present a grouped feature update module (GFUM) to reduce computational complexity by alternately updating high-level and low-level features within each group. Furthermore, the broadcast context module (CB) is introduced to obtain richer context information. It further enhances the ability to detect small objects. Extensive experiments conducted on three benchmarks, i.e. Remote Sensing Object Detection (RSOD), NWPU VHR-10 and PASCAL VOC2007, achieving 95.8%, 94.3% and 85.2% in mean average precision (mAP), respectively. Compared to DINO, the number of parameters is reduced by 10.4%, and the floating point operations (FLOPs) are reduced by 22.7%. The experimental results demonstrate the efficacy of ELFT in small object detection tasks, while maintaining an attractive level of computational complexity.

Turkish adaptation of a new scale for measuring transactional distance between students and the learning technology

Transactional distance is a key construct for understanding the quality of interactions in distance education. This study aimed to adapt the New Scale for Measuring Transactional Distance Between Students and Technology to Turkish culture. The adaptation process followed the 18-step guideline by the International Test Commission (ITC). The scale was applied to two study groups: Study Group 1 (n = 46) for linguistic equivalence and preliminary analysis, and Study Group 2 (n = 2402) for validation. Confirmatory factor analysis confirmed the four-factor structure with acceptable fit indices (CFI = 0.964, TLI = 0.947, RMSEA = 0.079, SRMR = 0.049). The reliability coefficients (Cronbach’s α) ranged from 0.702 to 0.939 across dimensions. Discriminant validity was supported through HTMT2 values, all below 0.90. The results suggest that the adapted scale is a valid and reliable tool for measuring transactional distance in Turkish distance education contexts.

A refined guide for aging muskoxen (Ovibos moschatus) based on mandibular examination

Accurately aging wildlife is essential for effective management and conservation efforts because it supports the estimation of demographic parameters used to model population dynamics and determine harvest quotas. Currently, accurately aging muskoxen is limited by the lack of validated and standardized protocols specific to this species. We investigated three methods for aging muskoxen: tooth eruption pattern, mandible morphometrics, and cementum annuli analysis (CAA). We examined 260 mandibles from community-harvested muskoxen with known harvest dates from Nunavut and the Northwest Territories and radiographed 89 of these mandibles with erupting teeth to track eruption stages. From these data, we developed a key to estimate age of muskoxen from newborn until all permanent teeth have completely erupted (60 months). Next, we assessed the relationship between muskox mandible morphometrics and age using 178 archived mandibles from Banks Island, Northwest Territories. Caudal mandible length was the strongest predictor of age in months (Adjusted R2 = 0.917) up to 5 years old (60 months), after which growth was negligible. This regression model included linear, quadratic, and interaction terms for caudal mandible length with sex. To evaluate accuracy of CAA for aging, we compared cementum results from incisors of 14 captive muskoxen to their known age using linear regression (Adjusted R2 = 0.847). We applied this model, fitted to captive muskox data, to predict the age of 32 community harvested adult muskoxen over 60 months old (5 yo+) using cementum age results. While there was a tendency to underestimate age, this method provided a more informative estimate than classifying all animals as adults once all teeth have erupted. Integrating these methods, we developed a decision tree to guide aging based on the putative age class and available sample type. This framework improves age estimation accuracy for harvested muskoxen, supporting population models and enabling more effective management and conservation.

Spatial structure evolution and ecosystem service relationship changes in urban-fringe-rural areas of megacities: Evidence from Suzhou, China

The evolution of urban-fringe-rural structures profoundly impacts ecosystem services (ESs). However, the way in which trade-offs and synergies in ESs respond to changes in regional spatial structures has rarely been discussed. This knowledge gap hinders the development of spatially explicit strategies to mitigate ecological degradation while accommodating urban growth, ultimately perpetuating unsustainable landscape management practices characterized by reactive rather than preventive interventions. Such critical disconnect between structural dynamics and ES feedbacks has emerged as a major bottleneck to operationalizing landscape sustainability in metropolitan regions. This study selected Suzhou—a typical megacity in China—as an example to conduct an empirical study. The urban, urban fringe, and rural areas were firstly identified in 2010 and 2022 using Deep Neural Network (DNN) based on multi-source geographical data. Then, seven typical ESs were assessed using multiple models, and their interactions were examined through correlation analysis, coupling coordination degree model, and a self-organizing feature mapping network approach. At last, this study highlighted the complex responses of ESs relationships to dramatically changing spatial structure of urban-fringe-rural areas and proposed landscape management strategies. The findings include the following: (1) from 2010 to 2022, the spatial structure of urban-fringe-rural areas in Suzhou changed considerably, with 69.04% rural areas transformed into fringe areas, and 50.83% fringe areas developed into urban areas; (2) based on transition process, the region was further divided into urban maintenance, urban expansion, fringe maintenance, fringe expansion, and rural retention areas. Most of the mean value of ESs showed a gradient increasing differences along urban-fringe-rural, while the greatest decrease occurs in fringe expansion and urban expansion areas; and (3) interactions for changes in ES pairs also more closely linked in these two regions, with synergies dominating. The coupled coordination index among multiple ESs declines significantly in these areas, degrading from key coordination to key or mild trade-offs bundles. The results show ES interactions exhibit significant spatial variability under the evolution of metropolitan spatial structure, thus innovatively proposing integration of ESs synergies into urban-fringe-rural development framework to support overall landscape sustainability.

Sectoral sensitivity of the Kuwait stock market to a dual shock

This study examines Kuwait’s stock market sectors’ response to the 2020 dual shock of the COVID-19 pandemic and the oil price war between Russia and Saudi Arabia. Traditional cointegration and causality models support the analysis alongside the frequency domain to assess short-run dynamics. The research sample integrates daily data from December 31, 2015, to February 23, 2022. The core research findings indicate that the consumer services, industrials, and basic materials sectors were the most affected by fluctuations in oil prices, highlighting the vulnerability of Kuwait’s oil-dependent economy. Furthermore, the results from the econometric modelling reveal a significant long-run relationship with the West Texas Intermediate index (WTI) and short-run dynamics with the Brent, OPEC and Dubai oil benchmarks. At the global level, monitoring the impact of shocks in commodity-driven economies can help in designing policies that minimise the effects of systemic risks, affecting energy supply chains and inflation stability. This research advances the field by providing a multi-methodological analysis integrating traditional and efficient econometric models to assess sectoral sensitivities in the context of Kuwait, which is an understudied oil-exporting economy. The research findings offer valuable insights for investors and policy-makers in managing risks associated with oil price fluctuations during times of enhanced uncertainty.

Linking distributed leadership to teachers’ innovation: Chain mediating roles of commitment and collaboration in Chinese schools

Teachers’ innovative pedagogical practices are central to enhancing professional development and improving instructional effectiveness. However, empirical studies investigating the association between distributed leadership and teachers’ innovative practices were limited in the Chinese educational context. This study examines the associations between distributed leadership and teachers’ innovative practices, with teacher commitment and teacher collaboration considered as potential mediators. Through structural equation modeling analysis of the Teaching and Learning International Survey (TALIS, n = 3,772) 2018 data, the findings indicate that distributed leadership is significantly linked to teachers’ innovative practices. Furthermore, teacher commitment and teacher collaboration exert a significant mediating effect between distributed leadership and teachers’ innovative practices, demonstrating a chained mediation effect. The findings underscore the necessity of highlighting the need for decentralized leadership, enhanced teacher commitment, and structured collaboration to support educational innovation. This study contributes to the theoretical understanding of distributed leadership and provides actionable strategies for enhancing teacher professionalism and supporting educational reform initiatives.

Estimating affective polarization on a social network

Concerns about polarization and hate speech on social media are widespread. Affective polarization, i.e., hostility among partisans, is crucial in this regard as it links political disagreements to hostile language online. However, only a few methods are available to measure how affectively polarized an online debate is, and the existing approaches do not investigate jointly two defining features of affective polarization: hostility and social distance. To address this methodological gap, we propose a network-based measure of affective polarization that combines both aspects – which allows them to be studied independently. We show that our measure accurately captures the relation between the level of disagreement and the hostility expressed towards others (affective component) and whom individuals choose to interact with or avoid (social distance component). Applying our measure to a large-scale Twitter data set on COVID-19, we find that affective polarization was low in February 2020 and increased to high levels as more users joined the Twitter discussion in the following months.

Effects of different lower limb movement velocities on trunk muscle activity and dynamic stability in the Sahrmann core stability test

Core stability is essential for both performance and rehabilitation, yet standardized movement velocity criteria for the Sahrmann Core Stability Test (SCST) remain undefined. This study aimed to analyze how lower limb movement velocity affects trunk muscle activation patterns across movement phases and sides during SCST Level 3 in a supine position. SCST Level 3 was selected because it challenges both deep and superficial muscles without being too difficult for healthy participants. Sixteen healthy adult males, recreationally active but not involved in a training program at the time of recruitment, performed leg-lowering movements at three velocities: slow (5 s), medium (3 s), and fast (1 s). Surface electromyography recorded activation in the internal oblique/transversus abdominis (IO/TrA), external oblique (EO), and rectus abdominis (RA). Dynamic stability was simultaneously monitored using an inflatable pressure-biofeedback unit positioned at L4–L5 and inflated to 40 mmHg; participants maintained the pressure within ±10 mmHg throughout each trial. Muscle activity was analyzed across five phases: 500 ms and 250 ms before movement (pre-500 ms, pre-250 ms), and three post-movement phases (post 1, post 2, post 3). A three-way repeated-measures ANOVA (factors: velocity, movement phase, and laterality) revealed significant velocity-phase interactions for all muscles (p < 0.05; IO/TrA: η2 = 0.170; EO: η2 = 0.198; RA: η2 = 0.153), indicating that velocity affected the temporal activation pattern, particularly during the middle phase of the movement. Faster movements led to a rapid increase in IO/TrA, EO, and RA activity at post 2, whereas slower movements showed a gradual increase. Activation levels converged across conditions by the final phase, with no significant differences between deep and superficial muscles or between sides. These findings suggest SCST Level 3 consistently elicits high trunk muscle activation regardless of velocity. However, faster movements may provide valuable insights into trunk stabilization during mid-movement for clinical evaluations.

Prediction of biological age using machine learning

In response to Taiwan’s rapidly aging population and the rising demand for personalized health care, accurately assessing individual physiological aging has become an essential area of study. This research utilizes health examination data to propose a machine learning-based biological age prediction model that quantifies physiological age through residual life estimation. The model leverages LightGBM, which shows an 11.40% improvement in predictive performance (R-squared) compared to the XGBoost model. In the experiments, the use of MICE imputation for missing data significantly enhanced prediction accuracy, resulting in a 23.35% improvement in predictive performance. Kaplan-Meier (K-M) estimator survival analysis revealed that the model effectively differentiates between groups with varying health levels, underscoring the validity of biological age as a health status indicator. Additionally, the model identified the top ten biomarkers most influential in aging for both men and women, with a 69.23% overlap with Taiwan’s leading causes of death and previously identified top health-impact factors, further validating its practical relevance. Through multidimensional health recommendations based on SHAP and PCC interpretations, if the health recommendations provided by the model are implemented, 64.58% of individuals could potentially extend their life expectancy. This study provides new methodological support and data backing for precision health interventions and life extension.

Fatigue and neuromuscular function in long COVID: A one-year follow-up study

Background Long COVID has emerged as a significant complication of SARS-CoV-2 infection. However, the long-term neuromuscular consequences of this condition, particularly one-year post-infection, remain poorly understood. This study aimed to determine the mechanisms of fatigue by comparing perceived fatigue, objective fatigability, functionality, muscle architecture, and electrical neuromuscular function in participants who had suffered severe or moderate COVID‑19 one-year post-infection with a healthy control group. Methods and Findings This longitudinal observational study followed participants for one-year. The assessments were conducted at the Laboratory of Muscle and Tendon Plasticity at the University of Brasília, Brazil. Participants who had suffered moderate or severe SARS-CoV-2 infection were compared to a control group. A baseline assessment was initially conducted (21–30 days post-symptoms onset or post-hospital discharge), followed by second (31–90), third (91–180), and fourth (181–360) assessments. Perceived fatigue, objective fatigability, functionality, muscle architecture, and electrical neuromuscular function were analyzed. The study included 30 controls (46.53 [42.10–51.43] years; 13 men [43.33%]), 22 moderate cases (38.27 [33.96–43.13] years; 10 men [45.45%]), and 18 severe cases (50.83 [45.19–57.18] years; eight men [44.44%]). Severe participants exhibited higher perceived fatigue in all assessments than the control group and at baseline and in assessment 4 compared to moderate cases, in addition to a lower torque and torque-time integral in all assessments of objective fatigability analysis compared to the other groups. The severe group also demonstrated reduced functionality, impaired muscle architecture (characterized by increased echogenicity), and higher chronaxie values in the electrical neuromuscular function assessment. Participants with moderate COVID‑19 exhibited alterations in perceived fatigue, reduced torque, and lower TTI, electrical neuromuscular function, and muscle architecture, particularly at baseline. Conclusions Severe participants continued to experience significant perceived fatigue even one-year post-infection, suggesting a slower recovery trajectory, that contributed to increased fatigability throughout the follow-up period. These results emphasize the role of musculoskeletal and neural mechanisms in post-COVID‑19 fatigue, highlighting the need for targeted, mechanism-based rehabilitation strategies. Trial registration NCT04961255

Remembering past emotions: How emotion expressions are linked to memory reappraisal

This paper examined how older adults experience the past and present valence of emotional memories and its relation to their multimodal expressions. As older adults can become more biased towards positive emotions and minimize negative emotions, a valence difference in the reappraisal of memories can exist. A study was conducted with older adults (N = 17) who self-reported their past and present valence of emotional memories. Data were collected between July and November 2018. Using automatic recognition technology, facial, acoustic, and lexical expressions were analysed during memory reminiscence. Results showed that differences exist between the past and present valence. Due to the reappraisal of the memory, a shift in valence occurred as a strong fading effect was seen in the sad memories, while a slight positive bias was also present as the older adults were more biased towards positive emotions in happy memories. Different modalities were more relevant for past than for present valence. These findings show that distinguishing between past and present valence is important when studying emotions in older adults’ memories. Older adults are more biased towards positive emotions and minimize negative emotions. Finally, automatic emotion recognition technology can benefit from these findings to assist older adults in the future.

Biomechanical tradeoffs in stroller running: Reduced vertical impact loading and increased torsional injury risk

This study evaluated how running with a stroller influences biomechanical parameters commonly associated with injury risk. Specifically, the study investigated changes in vertical and torsional loading and tibial acceleration in runners pushing a jogging stroller compared to running without one. Thirty-eight healthy adult runners participated in trials where they ran over a force plate with and without a stroller. Key measurements included vertical and anterior-posterior ground reaction forces, tibial accelerations, and free moments. The findings demonstrated significant reductions in vertical loading metrics, including vertical impact peak, vertical instantaneous and average loading rates, and vertical impulse, by 8–17% when running with a stroller. These reductions suggest a decreased risk of overuse injuries commonly associated with vertical forces. Conversely, torsional loading parameters, such as peak free moment and free moment impulse, increased significantly, with some measures rising by more than 400%. This increase in torsional loading indicates an elevated risk of stress-related injuries, particularly tibial stress fractures. In addition, tibial acceleration decreased slightly, though to a lesser extent than vertical loading metrics. These results highlight a biomechanical tradeoff when running with a stroller. While the reduction in vertical loading may mitigate the risk of bone stress and overuse injuries, the simultaneous increase in torsional loading could heighten the likelihood of torsional stress injuries. The study underscores the need for further research to explore mitigation strategies, such as optimized stroller designs or alternative pushing techniques, to balance these risks. The findings contribute valuable insights for runners, coaches, and stroller manufacturers aiming to promote safer running practices for caregivers using jogging strollers.

A Bayesian spatio-temporal model of variation in homicide rates for El Salvador

Most violence in El Salvador has been attributed to gangs. After being one of the most violent societies worldwide, in 2023, the country’s homicide rate reached a low 2.4 per 100,000 population. The decline is the outcome of a security plan by government that has attacked gangs directly under a state of exception. Nearly 80,000 gang members have been incarcerated. For the past 30 years or so, Salvadoran authorities and politicians alike, have participated in negotiations with gangs to reduce violence and gain electoral support in exchange for benefits for their members. This research studies violence as the outcome from the activities by gangs, politicians and governments, their interactions, and communities intervening in the realization of these interactions. As most data required for explaining these processes is either inexistent or difficult to access, a hierarchical Bayesian model was implemented for the spatio-temporal evolution of homicide with random effects that account for omitted variables at the level of local areas and time periods. The results support the view that unobserved covariates related to the district patterns of homicide have evolved over time. Two cycles appear in the evolution of homicide over the period under study, one from 2003 through to 2012, and another starting in 2013 and still going on at the time of writing. This finding reinforces the view that timing of government-gang negotiations drove the behavior of homicide rates in El Salvador during 2002–2021 together with the growth and expansion of gangs as seen from clustering of high-risk districts over time. The incarceration of scores of members and collaborators has both incapacitated gangs as key producers of violence, and deterred other forms of crimes. As a next step, the government should build collective efficacy, in particular among disadvantaged communities, to restrain the formation of gang-like groups in the time to come.

Regional disparities in the distribution of public and private healthcare facilities in South Korea

Unequal healthcare facility distributions remain a critical concern in South Korea, where medical resources are disproportionately concentrated in metropolitan areas, leading to disparities in the availability of services between regions. These spatial disparities are also influenced by broader contextual factors such as socioeconomic inequality and transportation infrastructure, which interact with access to care. Focusing on the spacial dimension of these disparities, this study investigates the spatial distribution of public and private medical facilities by analyzing their allocation patterns to identify underlying principles associated with these distributional trends, using a cost optimization approach rooted in statistical physics. By employing concentration index (CI), we quantify regional inequalities in general hospitals and clinics. Additionally, using scaling exponents, we aim to understand the principles underlying the distributional differences between public and private healthcare facilities and examine whether these differences vary across medical specialties. Our findings reveal that private hospitals and clinics exhibit stronger correlations with population density, reflecting commercial-like allocation patterns, while public institutions demonstrate more even spatial distributions. However, some essential medical specialties, including emergency medicine, pediatrics, and obstetrics and gynecology, are sparsely distributed in public sectors. Furthermore, temporal scaling exponent analysis suggests that private medical institutions are gradually intensifying their commercial characteristics, potentially exacerbating future disparities. These results provide a comprehensive understanding of the distribution patterns of healthcare facilities across regions and establishment types, while uncovering the underlying principles that shape these patterns based on medical specialties, providing valuable insights to inform future resource allocation strategies in the health sector.

Exploring risk and protective factors which distinguish suicidal and self-harm behaviours from suicidal and self-harm ideation in young people: A systematic review

Background Self-harm and suicidal thoughts and behaviours among young people are significant global public health concerns. Although most young people with thoughts of self-harm or suicide do not act on their thoughts, it is important to identify factors that distinguish thoughts of self-harm and suicide from behaviours. To date, there are no reviews distinguishing self-harm and suicidal behaviours from thoughts of self-harm and suicide in young people or that have synthesised factors distinguishing self-harm behaviours from self-harm ideation. The current review addresses these gaps in the literature. Methods We systematically searched: CINAHL, Embase, Medline, PsycINFO, Psychology and Behavioural Sciences Collection, and Web of Science Core Collection for articles published between 2011 and April 2024. Ninety-nine studies met inclusion criteria, with 92 articles examining risk and protective factors that distinguished suicide attempts from suicidal ideation and seven articles examining factors that distinguished self-harm behaviours from self-harm ideation. Using a narrative synthesis approach, studies were grouped by their outcome variable (e.g., self-harm or suicide) and then by risk and protective factors. Results While findings were inconsistent, the presence of non-suicidal self-injury, physical, emotional, or sexual abuse, violence, and family factors (e.g., family conflict) distinguished suicidal attempts from suicidal ideation. By contrast, the presence of parental factors (e.g., parental connectedness) and greater academic achievement were protective and distinguished suicidal ideation from suicide attempts. Being female, exposure to self-harm/suicide, and impulsivity distinguished self-harm behaviours from self-harm ideation. There was no evidence of protective factors that distinguished self-harm behaviours from self-harm ideation. Conclusions The current review highlights important risk and protective factors that distinguish suicidal and self-harm behaviours from suicidal and self-harm ideation in young people. Our review has important implications for intervention and prevention efforts as identifying key risk and protective factors can improve risk assessment for young people experiencing thoughts of self-harm and suicide and enable more targeted interventions.

Development of a proposed solutions box for improving nutritional behaviors of female school students: Protocol of a mixed-method study

Aim Adolescent girls in Iran face significant challenges related to poor nutritional behaviors, such as inadequate fruit and vegetable consumption, high intake of processed foods, and irregular meal patterns. These behaviors contribute to long-term health risks, including obesity and micronutrient deficiencies. ‘The aim of this study is to develop an evidence-based, context-specific ‘solution box’ to improve the nutritional behaviors of female school students in Khuzestan province. Design A mixed-method study using a sequential exploratory design (Creswell model). Methods This research will be conducted in three distinct stages. Stage 1: Qualitative Phase –Part A: A qualitative study using directed content analysis based on Social Cognitive Theory will be conducted. This phase will involve focus groups and interviews with female students, parents, and stakeholders to identify barriers to healthy eating and propose context-specific solutions. Part B: The perceived solutions will be extracted and categorized based on their relevance and feasibility. Stage 2: Quantitative Phase – Part A: Extracted solutions will be prioritized using the nominal group technique. Part B: Prioritized solutions will be ranked using a pairwise comparison scoring method. Stage 3: Solutions Optimization – Part A: A matrix of solutions will be developed based on theoretical frameworks. Part B: An action plan for implementing the solutions will be created. Part C: The solutions will be visualized on a digital platform for accessibility and usability. Results Not applicable for the protocol study.