Abstract: Introduction: Management of fractures or osteomyelitis with accompanying soft-tissue damage in the lower extremities remains challenging. This study evaluated the effectiveness of a multidisciplinary orthoplastic approach in patients with severe bone and soft-tissue defects of traumatic or infectious origin. Particular attention was given to the role of radiological techniques—Doppler ultrasound and Angio-CT—in preoperative planning. Imaging provided essential information on the number, location, size, and branching patterns of perforators, supporting functional and aesthetic outcomes. Material and Methods: This retrospective case series included 27 patients treated between May 2021 and December 2023. Group A comprised 17 patients with acute post-traumatic bone and soft-tissue loss, and Group B included 10 patients with chronic osteomyelitis, fistula, and soft-tissue defects. Treatment involved multidisciplinary reconstruction using local and free flaps following radiological assessment. Bone stabilization was achieved with low-contact locking plates or intramedullary nails. In osteomyelitis cases, polymethyl methacrylate bone cement served as a filler and local antibiotic carrier. Functional outcomes were assessed using the Lower Limb Functional Index (LLFI). Osteomyelitis and open fractures were classified according to Cierny–Mader and Gustilo–Anderson systems. Fourteen patients undergoing propeller or ALT flaps received preoperative Doppler ultrasound and CT angiography. Results: Mean follow-up was 17.9 months in Group A and 24.4 months in Group B. Doppler ultrasound findings were consistent with Angio CT results, with no false-positive perforators confirmed intraoperatively. Ultrasound additionally provided information on flow velocity and vessel diameter, supporting surgical planning. No infection recurrence was observed, and bone union was achieved in all patients. LLFI scores improved significantly (p<0.05), and all patients reported satisfaction with outcomes. Conclusion: A collaborative orthoplastic approach ensures favorable outcomes in complex lower limb reconstruction. Radiological assessment, particularly Doppler ultrasound, plays a key role in flap planning and may serve as a reliable standalone preoperative tool. Combined radical debridement, stable fixation, flap reconstruction, and local antibiotic therapy effectively manage complex orthoplastic conditions.

Abstract:

The roots of Acridocarpus smeathmannii were identified as a natural source of the benzyl-terpene 2-(5-isopropyl-4-methoxy-2-methylbenzyl)phenol (FAH-01, chamanen), which was isolated and structurally characterized by chromatographic and spectroscopic methods, including two-dimensional NMR analysis. Functionally, FAH-01 exerted pronounced inhibitory effects on human prostate smooth muscle contractility. In organ bath experiments, FAH-01 reduced noradrenaline-induced contractions by up to 72% and phenylephrine-induced contractions by up to 63%, without affecting agonist potency (pEC₅₀). During electrical field stimulation (2 - 32 Hz), FAH-01 significantly suppressed neurogenic contractile responses, indicating interference with adrenergic and nerve-mediated signaling pathways. Beyond smooth muscle modulation, FAH-01 showed antioxidant activity in the DPPH radical-scavenging assay and produced toxicity in the Artemia salina cysts. Collectively, these findings identify FAH-01 as a bioactive natural product with potent inhibitory effects on adrenergic and neurogenic contraction in human prostate smooth muscle, supporting its therapeutic potential in conditions associated with increased smooth muscle tone. Further preclinical studies to elucidate its mechanisms of action, safety profile, and in vivo efficacy are essential.

Abstract: The positive impact of strategic self-talk on performance in various sports tasks has been well-documented. Contemporary research has therefore started to explore the potential mechanisms through which self-talk affects performance. The purpose of the present study was to examine aspects of the attentional mechanisms underlying performance on a golf putting task, endorsing a psychophysiological perspective through the assessment of Heart-Rate Variability. Participants were 40 male sport science students, with no prior experience in golf putting, who were randomly assigned to control and experimental groups. The experiment was completed over four sessions, including baseline assessment, two training sessions, and final assessment. Participants of both groups followed the same training protocol, with the experimental group practicing strategic self-talk training and developing personal self-talk plans for the final assessment. Performance and HRV were recorded during the baseline and final assessments. Repeated measures analysis of variance showed that the experimental group showed greater performance improvement from baseline to final assessment and superior performance compared to the control group at the final assessment. Analysis of the HRV data provided indications that the experimental group showed different patterns of RMSSD activation at the final assessment, showing a greater activation of the para-sympathetic nervous system, in particular so, towards the latest stages of the golf-putting task. The findings are in line with an attentional interpretation of self-talk effectiveness through a more relaxed / less effortful processing, showing that when ego depletion was likely induced, the use of strategic self-talk mitigated any adverse effects.

Abstract: Objective. To describe sociodemographic and clinical characteristics, as well as evaluate knowledge about the disease in patients with acute myocardial infarction (AMI) for the first time in a hospital of high complexity in Bogotá, Colombia. Methods. A cross-sectional descriptive study was conducted in adults aged 18 and over diagnosed with a first episode of an AMI, in which data were collected through structured surveys that evaluated the cardiovascular risk profile, recognition of symptoms, life habits and knowledge of symptoms. Results. 61 patients, 73.33% men, with an average age of 57 years (SD±13) were included. The majority had elementary education as their highest level (34.43%), and more than half reported monthly income of between 1 and 2 million pesos (COP$) (55.74%). Hypertension (68.85%), diabetes mellitus (40%), and smoking (52.46%) were the most prevalent risk factors. 72.13% underwent coronary angioplasty, 45.90% were able to define the concept of AMI, being more frequent among those with higher educational levels (85.71%). Recognition of critical symptoms was limited, 54.10% identified chest pain, 49.2% dyspnea and only 37.70% syncope. 80.30% recognized that a healthy diet is preventive, 71.10% were physically inactive. In addition, 41% reported self-medication and 59% delayed seeking medical attention. Conclusions. Patients with the first episode of an AMI in our study population in Bogotá face a double burden: modifiable risk factors and wide knowledge gaps, especially among those who have lower educational level and low income. These findings high-light the need for public health interventions with an equity approach, including culturally appropriate educational programs and policies that encourage timely recognition and response to cardiovascular symptoms. Overcoming these barriers is crucial to improving outcomes in AMI and reducing preventable mortality in vulnerable populations.

Abstract: A control algorithm based on artificial neural networks was developed to regulate the hot-air drying temperature for carrot dehydration within an IoT-enabled cyber-physical system. The experimental setup employs an Arduino Mega 2560 equipped with AM2302, MLX90614, and SHT35 sensors, an HX711 load cell, and a WS68 anemometer, with cloud communication provided by an ESP8266 module for remote monitoring via Wi-Fi under an IoT framework. The neural controller, implemented using the Arduino Neurona li-brary, adjusts the dryer temperature in real time, ensuring thermal stability throughout operation. Three initial loads (2, 4, and 6 kg) were analyzed to obtain the drying kinetics and determine the thermal efficiency. In the dehydration experiments, the 2 kg load reached a final moisture content of 10% in 4.4 h, consuming 1,390 kJ with a thermal effi-ciency of 83%. The 4 kg load exhibited the best time–energy balance (6.6 h, 1,850.0 kJ, 88%), while the 6 kg load achieved the highest efficiency (8.1 h, 2,250.0 kJ, 91%). These results demonstrate the effectiveness of neural-network-based control implemented on low-cost microcontrollers to enhance thermal efficiency in food dehydration processes.

Abstract: Background: Accurate prediction of lip prominence changes following orthodontic treatment remains challenging because traditional profile analyses rely on isolated reference lines that do not account for combined nasal and chin morphology. The aesthetic triangle framework integrates these structures and may provide a more comprehensive evaluation of lip position. Methods: This correlative clinical study evaluated 82 orthodontic patients undergoing bimaxillary incisor repositioning. Lateral cephalograms and standardized profile photographs were obtained before and after treatment. Lip position was assessed relative to the aesthetic triangle boundaries, and dentoalveolar changes were quantified using standard incisor measurements. Lip thickness was analyzed as a potential modulating factor. Results: Orthodontic treatment produced significant incisor proclination and sagittal advancement. Within the aesthetic triangle framework, the lower lip demonstrated consistent anterior migration toward the central compartment following mandibular incisor advancement. In contrast, the upper lip showed minimal displacement despite substantial maxillary incisor movement. Mandibular incisor inclination significantly predicted lower lip migration, while maxillary incisor variables showed limited predictive value. Thinner lips exhibited greater positional responsiveness. Conclusions: The aesthetic triangle provides a clinically meaningful framework for interpreting orthodontic soft tissue changes as spatial migration rather than isolated linear measurements. Lower lip prominence responds predictably to dentoalveolar mechanics, whereas upper lip position is more constrained by soft tissue morphology.

Abstract: Pharmacomicrobiomics is the study of drug-microbiome interactions. In other words, it examines the dynamic relationship between the drug, the host, and the microbiome. This has become a rapidly evolving area in the realm of pharmacology and personalized medicine. Emerging evidence demonstrates that the gut microbiome can influence the pharmacodynamics and pharmacokinetics of drugs through various mechanisms while drugs can simultaneously alter microbial composition. Treatment approaches include regular targeted pharmaceutical therapies (eg: antibiotics, antidepressants), alternative treatment approaches (eg: CAM treatments- supplements, herbs), non-drug biological therapies, like ECT and TMS. The new term for this approach called Microbiome Based Medication Treatment, which has been seen as an alternative treatment approach and has been studied extensively in the last decade. This review article focuses on current knowledge on drug–microbiome interactions across multiple therapeutic systems, including cardiovascular, central nervous system, gastrointestinal, respiratory, endocrine, oncologic, musculoskeletal, and anti-infective therapies. Furthermore, we highlight the various pathways by which microbes can alter the various mechanisms like drug absorption, bioavailability, efficacy, and incidence of adverse effects, along with the clinical implications of drug-induced dysbiosis.

Abstract: In healthcare organizations, ransomware threats are on the rise, resulting in a greater disruption of critical patient care operations, yet current cybersecurity approaches are mostly reactive rather than proactive. As part of this study, a systematic comparative evaluation of traditional machine learning versus deep learning methodologies on small-scale tabular cybersecurity datasets characteristic of healthcare security operations is conducted to address the critical knowledge gap regarding optimal algorithmic approaches for healthcare ransomware threat prediction. In this study, the Healthcare Ransomware Dataset is used, which contains 5,000 simulated attack records with no missing values across 16 attributes capturing organizational attributes, attack characteristics, and outcome metrics. This study evaluated eight machine learning algorithms (Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, Extra Trees, AdaBoost, Naive Bayes, K-Nearest Neighbors) and eight neural network architectures (Simple Deep Neural Network, Wide Deep Neural Network, Residual DNN, CNN_1D, LSTM, GRU, Ensemble DNN) across five performance dimensions: accuracy, precision, recall, F1-score, and ROC-AUC. According to the feature importance analysis, there is a strong correlation between organizational size (27.94%), recovery time (16.18%), and data restoration (12.06%) which account for 56% of the predictive power. A gradient boosting approach achieved 84.1% accuracy with 95.19% ROC-AUC and Simple DNN represented the best deep learning architecture with 85.2% accuracy and 95.44% ROC-AUC using only 12,355 parameters. As compared to machine learning (5.19%), deep learning demonstrated significantly lower performance variance (1.64% standard deviation). Across all sequence architectures (CNN_1D, LSTM, GRU), sequential architectures consistently underperformed by 2.6-4.5 percentage points, confirming the architectural mismatch for unordered tabular features. Based on the results of the analysis, Simple DNN provides the highest predictive accuracy (85.2%) for healthcare ransomware threat assessment. The complexity of an architecture beyond shallow dense networks yields diminishing returns without improving accuracy. In comparison to traditional machine learning ap-proaches, deep learning offers superior consistency across a variety of architectures. It is recommended that Simple DNN be de-ployed as the primary prediction model and Gradient Boosting be used as the fault-tolerant backup model. It is important to place a high priority on the eight top features that capture 90 percent of the predictive power of data. In the case of unordered tabular cybersecurity data, sequential architectures (CNNs, LSTMs, GRUs) should be avoided.

Abstract: Rural architectural heritage sites in Korea, including rice mills, breweries, and granaries, face an increasing risk of neglect, damage, and demolition. Because most of these structures lack recognition in formal heritage designation systems, their conservation and management are challenging. This study proposes a comprehensive evaluation framework for rural architectural heritage. Based on a literature review and expert consultations, we derived 18 evaluation indicators grouped into six value criteria: historical, architectural/artistic, social/cultural, landscape, economic, and utilitarian values. The Analytic Hierarchy Process (AHP) was employed to determine the relative importance and priority of these indicators. The results indicate that historical value had the highest weight among the six criteria, followed by architectural/artistic and social/cultural values. Among the 18 indicators, “representativeness of the period” ranked highest, followed by “rarity,” “historicity,” and "architectural excellence." However, the indicators associated with economic and utilitarian values received relatively low weights. The framework validated by applying it to 17 rural heritage sites in Buyeo County, a representative rural region in Korea. This study presents a systematic and value-based evaluation framework that reflects the regional and industrial characteristics of rural architectural heritage and provides theoretical and practical implications for its conservation and adaptive reuse.

Abstract: As Germany advances toward its 2030 energy transition targets, large‑scale battery systems are widely promoted as a key instrument to manage the volatility of wind and solar power. This paper develops a data‑driven simulation framework to quantify how much grid‑scale batteries can realistically contribute to Germany’s power system in 2024 and in a 2030 expansion scenario, in terms of backup energy reduction, system costs, and carbon footprint. The analysis uses 15‑minute data for generation and load from 2022–2025 and constructs future scenarios by upscaling photovoltaics, onshore wind, offshore wind, and demand in line with the Study "Climate‑neutral Germany 2045". Volatility is decomposed into a slowly varying component, covered by backup plants, and a normalized residual load that feeds the batteries; for each battery capacity, an optimal moving‑average window is determined to maximize annual battery output and minimize backup energy. Model validation against 2024 pumped‑storage operation shows that the framework reproduces observed annual storage output well. Across both 2024 and 2030, battery benefits scale approximately with the logarithm of installed capacity, while costs and battery‑related emissions increase linearly. Under optimistic assumptions, about 300 GWh of batteries can supply only 4–6% of annual demand and cannot replace firm backup capacity, implying that batteries are valuable for short‑term balancing but have limited potential for seasonal adequacy.

Abstract: Parvoviruses are small, non-enveloped DNA viruses, several of which have been adapted as vectors for gene therapy. Adeno-associated virus (AAV) is clinically established but constrained by limited genome capacity and pre-existing immunity. Human bocaviruses (HBoVs) possess larger packaging potential and airway tropism, motivating exploration of AAV-HBoV hybrid architectures. We modeled a chimeric construct (AAV-HB3) in which the α-helix and βH/βI strand of AAVGo.1 were replaced with the corresponding regions from HBoV3. AlphaFold2 predictions (pLDDT > 80, pTM > 0.75) confirmed reten-tion of the β-barrel scaffold, and RoseTTAFold2 refinement produced energetically stable conformations. 100 ns molecular-dynamics simulations showed distinct dynamic profiles: the AAVGo.1 VP1 control remained conformationally rigid (RMSD ≈ 0.03 nm), whereas AAV-HB3 exhibited increased flexibility at VR-VIII, VR-IV, and the HI loop, with loop dis-placements of ~43–54 Å localized near the three-fold and five-fold symmetry axes. These structure-based analyses define how cross-genus substitutions redistribute local flexibility within the parvoviral capsid, providing a predictive framework for engineering next-generation vectors with assessing structural tolerance relevant to capsid.

Abstract: The resummation of superfactorially divergent series represents a significant computational challenge in mathematical physics. In the present paper the resummation of a specific class of Stieltjes series characterized by a moment sequence growing as $(2n)!$ will be addressed. Despite the fact that Carleman's condition is satisfied for these series, the practical utility of Padé approximants in resumming them is severely compromised by an extremely slow convergence rate. Weniger's $\delta$-transformation is proposed as a superior resummation tool. Some recently found results on the converging factors of superfactorially divergent Stieltjes series are here used to derive an exact integral representation of the $\delta$ truncation error, which allows for a formal proof of convergence and an analytical asymptotic estimate of the corresponding convergence rate. Numerical experiments are carried out to validate our theoretical findings, confirming that the $\delta$ transformation offers a robust and computationally efficient framework for decoding this class of wildly divergent expansions.

Abstract: Background: The gut microbiota and plasma metabolites have been shown to contribute to the etiology of Post-Traumatic Stress Disorder (PTSD). The relationship between the gut microbiome and plasma metabolome in PTSD is poorly understood. This study aims To integrate the gut microbiome data and plasma metabolome data to elucidate microbial–metabolite associations specific for PTSD in a mouse model. Methods: A PTSD mouse model was induced by Single Prolonged Stress and Electric Foot Shock (SPS&amp;S). We sequenced gut microbiota composition by 16S rRNA gene sequencing and used Ultra-high-performance liquid chromatography–tandem mass spectrometry (UHPLC-MS/MS) for plasma metabolomic profiling, to explore the association Between gut microbiota and plasma metabolites in mice with PTSD. Results: UHPLC-MS/MS analysis showed significant differences in the abundance of multiple plasma metabolites, including L-arginine, palmitic acid, oleic acid, uridine, in mice with PTSD versus control mice. Pathway enrichment analysis further revealed that these differentially abundant metabolites were all enriched in pathways such as arginine biosynthesis, unsaturated fatty acid biosynthesis, pyrimidine metabolism and glycerophospholipid metabolism. Meanwhile, 16S rRNA sequencing revealed differences in the composition and diversity of the gut microbes in PTSD mice and control mice. For example, the relative abundance of the Muribaculaceae and Akkermansia genera decreased significantly in the PTSD group, while relative abundance of Lachnospiraceae NK4A136_group increased significantly in the PTSD group. Further correlation analysis showed strong correlation between the plasma metabolome and gut microbiome. Conclusions: These findings confirm the association between the gut microbiome and the plasma metabolome in PTSD and point to these specific taxa and their connected metabolites as potential biomarkers and treatment targets in PTSD.

Abstract: LAIB formulations have emerged as an alternative pharmacological approach for opioid use disorder, offering potential benefits extending beyond clinical stabilisation. Narrative medicine provides a unique approach to understand patients’ perspectives and experiences with sublingual buprenorphine and LAIB, and to examine how they impact the emotional, social, and motivational dimensions of recovery. Fourteen patients across different Italian Addiction Services were interviewed. Narratives were analysed by thematic content across eight domains: dependence on daily treatment regimen, emotional impact, self-perception, determination to change, quality of life, craving and withdrawal symptoms, treatment adherence, social burden, and therapeutic relationship. Statements were categorised by valence; experiential patterns were qualitatively analysed. Sublingual buprenorphine, although effective, was associated with reduced autonomy, symptom control, and difficulties in balancing treatment, work and life. These aspects were correlated with worse adherence. The stigma and burden of daily intake can reduce motivation and hinder identity reconstruction. Transitioning to LAIB resulted in improved self-autonomy, emotional balance, symptom control, self-esteem, and reduced daily and psychological burden, craving and stigma, facilitating social reintegration, and strengthening the therapeutic relationship. The results emphasise how crucial it is to include both experiential and narrative elements in clinical care, as this helps create more tailored, recovery-focused treatment pathways.

Abstract: Background/Objectives: Childhood overweight and obesity are on the rise in the United States with one in five children reported to be obese. This can lead to increased risk of di-et-related chronic diseases earlier in life. Reports conflict on whether fruit and vegetable programs are beneficial. The aim of this study was to look at the Real Food for Kids (RFFK) Fresh Food Explorers Program, a 8-week program that aimed at improving preschoolers' views of vegetables in different categories. Methods: Preschool students completed visual surveys at the beginning and end of the program, asking about their views of vegetables from different categories. Following the pre-survey, teachers were provided weekly kits over eight weeks containing videos, lesson plans, activity sheets, and detailed hands-on activities. Weekly lessons focused on specific vegetables, including tastings, to reinforce lesson concepts and build positive exposure. Children were sent home with produce bags to share with their families, allowing the les-sons and exposure to continue at home and helping to address food insecurity in house-holds. . Survey results were documented and analyzed to determine the impact. Results: A total of 177 preschoolers from 10 classrooms in the Washington DC area showed an increase in positive views on vegetables in all categories except for tomatoes, which remained the same. Conclusion: Our findings show that the Fresh Food Explorers program may improve veg-etable consumption and views among preschool children and show that knowledge and repeat exposure are essential.

Abstract: Bell tests and Bell's theorem used to interpret the test results opened the door to quantum information processing, such as quantum computation and quantum communication. Based on the erroneous interpretation of the test results, quantum information processing contradicts a well-established mathematical fact in point-set topology. In this study, the feasibility of quantum computation and quantum communication is investigated. The findings are as follows. (a) Experimentally confirmed statistical predictions of quantum mechanics are not evidence of experimentally realized quantum information processing systems. (b) Physical carriers of quantum information coded by quantum bits (qubits) do not exist in the real world. (c) Einstein's ensemble interpretation of wave-function not only will eliminate inexplicable weirdness in quantum physics but also can help us see clearly none of quantum objects in the real world carry quantum information. The findings lead to an inevitable conclusion: Without carriers representing quantum information, physical implementations of quantum information processing systems are merely an unrealizable myth. Examples are given for illustrating the reported results. For readers who are unfamiliar with point-set topology, the examples may alleviate difficulty in understanding the results.

Abstract: Purpose: Physical activity levels tend to be lower in Parkinson’s disease (PD) than in healthy age-matched controls, largely because of substantial gait impairment. Remarkably, individuals with PD often retain a remarkable ability to cycle. However, so far, no evidence exists regarding the safety of patients when cycling outdoors in daily life. Therefore, the aim of this review is threefold: 1) to compare the mechanisms of cycling to those of walking, 2) to examine the symptomatic effects of stationary cycling, and 3) to highlight the challenges and opportunities to ensure safe outdoor cycling in PD.Major findings: Compared to walking, making cycling movements likely engages alternative neural circuits that largely bypass the most affected parts of the striatum in PD. Stationary cycling programs are therefore feasible and consistently improve physical fitness and motor symptom severity. Outdoor cycling may further increase physical activity levels, social participation and mobility in PD. However, its feasibility and safety are poorly studied, while there exist significant safety concerns linked primarily to poor balance, but also axial rotation, motor initiation and cognitive impairments. The present review discusses these challenges and considers which may be modifiable through future cycling-rehabilitation interventions.Conclusion: The relatively preserved ability to make cycling movements in PD makes outdoor cycling an attractive form of physical activity and promising strategy to enhance quality of life. However, further research is needed to assess outdoor cycling skills and inform rehabilitation interventions to ensure safe participation in a population for whom maintaining physical activity is essential.

Abstract: Environmental awareness is playing an increasingly important role in all segments of the world, with sustainability and recycling being key elements. The aim of the research is to examine the challenges companies face in terms of sustainability when implementing procurement and supply chain management processes related to interior design. The research focused on four main questions: how procurement and supply chain man-agement are reflected in construction processes, what challenges these processes face, and how they can influence the sustainable use of materials in architectural supply chains. The literature review was based on a systematic literature review using the PRISMA screening process and the PEO framework, utilizing the SCOPUS database and pro-cessing 70 scientific articles following the selection process. During the research, I also used the Critical Incident Technique (CIT), in which I asked interior designers about their positive and negative experiences with the procurement of sustainable materials and supply chain management processes. The methodology thus provided deeper insight into the decision-making processes of professionals, where sustainability conflicts with eco-nomic and operational realities. The qualitative research was supplemented by a ques-tionnaire survey, which aimed to assess sustainability, its prevalence, and professional obstacles. The results of the research show that this topic is a research gap, but the openness of professionals shows a positive trend. Companies face numerous challenges related to new technologies and environmental awareness in order to create or transform well-functioning supply chain management processes.

Abstract: Background/Objectives: The purpose of this study was (i) to determine the levels of interleukins in patients with COVID-19 admitted to the Intensive Care Unit (ICU), and (ii) to study their evolution over 72 hours, as well as (iii) to assess their relationship with morbidity and mortality. Methods: This was a prospective analytical study of critically ill patients with COVID-19 who were monitored from admission to three days of stay in the ICU. Demographics, comorbidities, clinical data and severity scores were recorded. IL-1β, IL-2, IL-6, IL-7, IL-8, IL-10 and antitumor necrosis factor alfa (TNF-α) were measured on the LuminexR 200™ Systems by an immunoassay determination method (ELISA). Clinical variables were collected and analytical variables were measured on the first and third day of admission. Results: In this study a decrease in IL-10 and IL-1β was observed, and an increase in TNF-α on the third day of ICU stay in critically ill patients with COVID-19. TNF-α, IL-10 and IL-8 are found to be higher in COVID-19 patients who die than in those who do not. In addition, we found an association of interleukins with clinical and laboratory variables. Conclusions: IL-10, TNF- α and IL-8 are useful for establishing a prognosis of morbidity and mortality in the first hours of ICU admission in patients with COVID-19. Interleukin levels (IL-10, IL-1β and TNF-α) evolve in the first three days of ICU admission towards a state of increased hyperinflammation that are related to clinical and analytical variables.

Abstract: This study addresses modeling requirements in real business scenarios where data distributions are complex and continuously evolving. It investigates a unified modeling problem under the coexistence of class imbalance and distribution shift. To tackle practical challenges such as scarce high-risk event samples, majority class dominance in learning, and decision bias amplified by environmental change, a collaborative modeling framework is developed. The framework introduces class structure awareness and distribution stability constraints within a shared representation space. This design enables the model to preserve discriminative capability while reducing dependence on a single training distribution. Structural bias caused by imbalance is alleviated through explicit modeling of class proportion differences. Interference from environmental change is suppressed through distribution consistency constraints during representation learning. As a result, model stability and consistency across different data conditions are improved. To examine controllability and robustness, the study conducts a systematic analysis of key hyperparameters, data quality degradation, and environmental perturbations on risk ranking performance. Special attention is given to sensitivity patterns induced by feature noise injection and varying missing value ratios. The results indicate that when class imbalance and distribution shift interact, collaborative modeling effectively mitigates performance degradation. The model maintains a stable risk recognition capability in complex business data. This provides a practically adaptive modeling approach for data-driven decision-making in high-risk scenarios.