Open Research Oklahoma (Oklahoma State Univ.)
Not a member yet
42035 research outputs found
Sort by
Tailored preventative care recommendations using electronic health records
No full textAccording to the U.S. Centers for Disease Control and Prevention (CDC), 90% of the nation’s $3.3 trillion annual healthcare expenditures are attributed to individuals with chronic and mental health conditions. Thus, preventing diseases is essential to improving public health and managing escalating healthcare costs. However, the preventive care clinical decision support (CDS) modules in most electronic health record (EHR) systems primarily rely on basic criteria such as age, gender, and screening intervals. This “one-size-fits-all” approach fails to provide personalized recommendations that consider patient-specific risk factors, such as family history, social history, ethnicity, and chronic conditions. This research develops an information system that analyze patient-specific data against the information extracted from preventive care guidelines to generate tailored recommendations along with justifications grounded in both the EHR data and preventive care guidelines.
Our system empowers patients by providing personalized insights into their potential health risks before issues arise. By analyzing factors such as family history, social background, ethnicity, chronic conditions, age, gender, and past medical history, it delivers tailored recommendations based on established guidelines. With this information, individuals can take a proactive approach to their health, engaging in informed discussions with their physicians to explore preventive measures. By promoting early intervention and personalized care, this system has the potential to reduce the burden of preventable diseases and improve long-term health outcomes.Management Science and Information System
Fentanyl administration in mice: A comparative study of method vapor self-administration vs intraperitoneal injection
No full textIntroduction: The synthetic opioid crisis, particularly involving fentanyl, has led to a dramatic surge in overdose deaths in the United States, with over 70,000 fatalities in 2021. Oklahoma reflects this trend, reporting 1,196 overdose deaths in 2022—a record rate of 30 per 100,000 residents. Fentanyl-related deaths in the state skyrocketed from 47 in 2019 to 300 in 2022. This crisis underscores the urgent need for advanced research into opioid use disorders and treatments. While the intravenous selfadministration model in rodents is the current gold standard for studying opioid addiction, it presents challenges such as catheter placement difficulties and maintenance issues. To overcome these obstacles, a vapor self-administration model for fentanyl delivery is being considered as a promising alternative, potentially enhancing research efficiency and providing new insights into addiction mechanisms and therapeutic approaches.Objectives: To determine if the pharmacodynamic of fentanyl via vapor self-administration was comparable to IP administration.Methods:96 naive mice were divided into 12 groupsFentanyl was administered through IP injections and vapor routes at equipotent dosesThe analgesic effect of fentanyl was evaluated using the hot-plate testConclusion: Hot-plate assays revealed equivalent analgesic effects in mice following fentanyl administration via intraperitoneal (IP) injection and passive vapor self-administration routes. The vapor self-administration paradigm represents a significant innovation in preclinical addiction research, offering enhanced translational potential for investigating substance use disorders
Exploring the characteristics of highly rated books
No full textUnderstanding the characteristics that contribute to five-star reviews is essential for authors and publishers aiming to optimize book development and marketing strategies. Despite the significance of this topic, limited data driven research exists on how book attributes influence reader ratings. This study addresses the research question: What characteristics lead to higher average book ratings? The goal is to identify key factors that drive higher ratings and provide actionable insights for positioning books to enhance reader satisfaction. The analysis methods utilized include exploratory data analysis, principal components analysis, and decision tree modeling. First, exploratory data analysis helps to identify relationships between characteristics of books and their ratings. Then, principal components analysis (PCA) is used to group genres as variable inputs for the model. Lastly, decision tree modeling assesses the predictive power and significance of book characteristics. Findings suggest that price, number of reviews, and specific genres are associated with higher ratings. This research provides valuable insights into reader preferences, enabling authors and publishers to tailor their book development strategies and marketing approaches. These findings can improve reader satisfaction, increase book popularity, and guide authors and publishers in refining their overall strategies for success in a competitive market.Marketing and International Busines
Evaluating data sharing statements in high yield psychiatry journals: A systematic review
No full textBackground: Psychiatric disorders affect approximately one in eight people worldwide, with economic impacts projected to reach $6 trillion by 2030. Effective data sharing is crucial for advancing psychiatric research, yet its practice remains inconsistent due to ethical and regulatory challenges. This review aimed to evaluate the prevalence and determinants of data sharing statements (DSS) in psychiatric research. We focused on how journal policies, study design, and other factors influence DSS inclusion in clinical trials and observational studies, examining trends and barriers to transparency and reproducibility.Methods: We systematically reviewed articles from the ten highest-impact psychiatric journals published between January 1, 2018, and December 31, 2023. We assessed DSS presence, tracked trends, and analyzed factors affecting DSS inclusion using hierarchical logistic regression. A thematic analysis categorized DSS content and evaluated the follow-through on data sharing promises.Results: Of 1,042 reviewed articles, 319 included DSS. The proportion of articles with DSS rose from 5% in 2017 to 56% in 2023. DSS inclusion varied significantly among journals; The Lancet Psychiatry and The British Journal of Psychiatry reported high inclusion rates, while the American Journal of Psychiatry had low rates. Clinical trials were more likely to include DSS than other study designs. Regression analysis identified higher journal impact factors, open access status, and study design as significant predictors of DSS presence. Thematic analysis highlighted prevalent themes like 'Gatekeeper Role' and 'Conditional Data Availability,' reflecting concerns about data privacy and accessibility.Conclusions: Despite a rising trend in DSS inclusion, substantial variability exists across journals and study designs. Journal policies and impact factors play a critical role in DSS presence. To enhance research transparency and reproducibility, standardizing data sharing guidelines, improving journal policies, and providing better researcher education are essential. Addressing ethical and legal concerns while promoting open science practices will improve research quality and psychiatric care outcomes
Social support and friendship quality across contexts: Exploring digital versus in-person relationships of emerging adults
No full textThe social lives of college students have expanded over the 21st century to include online spaces. Little is known about the importance of social relationships occurring online versus offline (i.e., in face-to-face situations). Prior research shows social support as generally strongly related to well-being. Social support is also found to be a motivator to use social media to connect with others. Intersecting social media use, well-being, and friendship quality provides insight into modern connections. In the present research, we hypothesized offline friendship quality will be more strongly related to wellbeing measures than online (i.e., connections through technology) friendship quality. We also hypothesized that men and women would differ in amount of offline and online social support and overall well-being. Men and women are also anticipated to differ in how social media use. Finally, we hypothesized participants to report more offline than online social interaction. Using Qualtrics to compile online questionnaires, we assessed online and offline social support, wellbeing (i.e., life satisfaction, loneliness, and perceived stress), and time spent interacting online and offline within the last two weeks. Using the Department of Psychology SONA systems, we recruited 237 participants (93 men, 149 women, 2 other). Results indicated partial support for the hypotheses. Social media usage was supported as a significant predictor for perceived stress, but not for loneliness and life satisfaction. Social support of a particular online friend was not related to well-being, but social support from a particular offline friend was. Unexpectedly, there were fewer gender differences than hypothesized. Usage of social media platforms showed the most variation between men and women. Women reported higher levels of stress than men. Lastly, we confirmed the hypothesis that women and men spend more time in offline than online social interactions.Psycholog
Re-routing: OpenOKState undergraduate student outreach
No full textThis presentation explores a collaborative outreach initiative between an open educational resources (OER) program and a university student council. The project focuses on empowering undergraduate students through OER advocacy, retention support, and inclusive academic engagement. Drawing on student leadership and institutional collaboration, the session highlights strategies for sustainable student involvement in open education.falseLibrar
Statistical effects on FEA-simulated stress and strain in feeding rodents when teeth are modeled and implanted as separate structures
No full textBackground: Finite element analysis (FEA) is used in paleontology to visualize the stress and strain distributions in extinct animals during biting and to make ecological inferences. Most previous work models teeth as physically and materially continuous with the cranium and mandible. When teeth are modeled separately the focus is usually on individual teeth or the mandible. We test the impacts of modeling the teeth as separate structures on the cranium, mandible, and teeth, with the first statistical comparison for mammals of stress and strain results in separate-tooth versus simplified continuous models.Methods: We used CT scans of Rattus norvegicus (ID: 55306) and Cavia porcellus (ID: 55304) from Morphosource. We produced three models: (1) a continuous bone-material model of the cranium and mandible with the teeth fused, and models with teeth as separate objects with; (2) enamel or; (3) dentine properties applied. We imported models into Strand7 for FEA. Forces from published literature were applied for the muscles of mastication: the superficial masseter, anterior and posterior deep masseter, temporalis, anterior, posterior, and infraorbital zygomaticomandibularis, and the internal and external pterygoid muscles. We applied elastic moduli of 17 GPa for bone, 16.9 GPa for dentine, and 83 GPa for enamel. By constraining respective teeth, we simulated a bilateral incisor bite to simulate gnawing, and the start of a unilateral left sided chewing motion. Simulated direct attachments connected the teeth to the cranium and mandible.Results: von Mises overall stress, and von Mises, tensile and compressive (first and third principal) strains, were sampled from twenty-two consistent mandible and tooth locations, and compared stress and strain between models using Kruskal-Wallis and Tukey-Kramer tests. Modeling the teeth as separate has a varying effect on the resulting stress in the two taxa. The rat showed more differences between the three models. The guinea pig model showed no significant difference in stress when modeling teeth as separate structures. The dentine separate teeth analysis resulted in both the guinea pig bites having overall higher strain compared to the bone and enamel models. The rat molar bite with teeth modeled as enamel showed high von Mises and tensile strain anterior to the left maxillary incisor, and the dorsal surface of the left mandibular ramus. With incisor bites, the highest strain occurred with teeth as dentine and lowest strain on the all-bone model.Conclusions: This is the first study in rodents that examines the statistical differences in FEA results when teeth are modeled as separate structures. We demonstrate that such modeling promises a more realistic distribution of stress and strain during both molar and incisor biting. Detaching the teeth from the jaw reveals where stress and strain transfer from the teeth to the skull. The difference in modeling teeth as enamel or dentine is expected, as enamel has lower strain values with its greater stiffness. The difference in results between the rat and guinea pig are potentially related to differences in diet, chewing motion, or phylogeny. This will be resolved in future research by including more rodents
Novel production methods in controlled and semi-controlled environments
No full textAs arable land decreases and the world population increases, advancements in controlled and semi-controlled environment agriculture are necessary to produce adequate amounts of food. Three such advancements are aquaponics, biochar, and arbuscular mycorrhiza (AM) fungi. For Experiment 1, two planting densities of two leafy greens and two fruiting crops were examined in aquaponic systems stocked with bluegill in a hoophouse in Stillwater, OK and grown year-round. Although the two fruiting crops did not produce significantly different fruit yields between the two densities, kale required more space to grow and lower planting density. Aquaponic systems were successfully run yearlong without significant environmental controls beyond water heaters in Oklahoma. For Experiment 2, eastern red cedar biochar was mixed into a peat-based media at four rates (0, 15, 30, and 45% of the total pot volume) and inoculated with five mycorrhizal sources (AMF-, two commercial AM fungi sources, and two extracted native spore sources, prairie and flower farm) to grow geraniums under greenhouse conditions. Biochar rates > 30% negatively impacted plant and AM fungal growth. Mycorrhizal fungi colonization did not mitigate the negative effects of biochar on geraniums, reducing plant vegetative and reproductive growth and increasing nutrient deficiencies. For Experiment 3, different types of biochar (eastern red cedar, leached eastern red cedar, pistachio shell, and hardwood) were added to geraniums at 0, 15, and 30% of the total pot volume. Commercial AM fungi (Mycobloom) was used to inoculate half the pots. Pistachio shell biochar at 15% did not negatively impact geraniums compared to the control, whereas pistachio shell biochar at 30% limited plant growth significantly. Although eastern red cedar biochar at 30% improved AM fungal colonization, AM fungi-plant symbiosis did not improve geranium growth. For Experiment 4, four native AM fungi sources (AMF-, flower farm, vegetable farm, and turf area) were compared to six species of plants (tomato, cucumber, zinnia, marigold, buffalo grass, and Kentucky bluegrass). No AM fungal inoculum was universally applicable to all species, and AM fungal inoculation did not significantly improve plant growth. Vegetable farm and flower farm inoculum had the greatest AM fungi colonization rate compared to AMF- and turf inoculum, but colonization did not significantly improve plant growth
Dynamics of the apparent contact angle on porous and dense cylindrical surfaces
No full textAn experimental study of the dynamics of the apparent contact angle, motivated by various applications, including dropwise condensation, heat and mass transfer, and flow separation, is presented. The study investigates the dynamics of the apparent contact angle and wetting behavior of droplets on both dense and porous hydrophobic tubular surfaces. Test liquids include distilled water, ethyl alcohol, Glycerol, and 35 g/L NaCl solution. Diagnostics consisted of high-speed backlight imaging to track the droplet geometry and contact angle evolution over time. A theoretical model was developed to investigate the temporal effect on the apparent contact angle due to evaporation. The study identifies three distinct stages of droplet base dynamics: expanding, constant, and shrinking. On porous polypropylene membrane, the apparent contact angle initially decreased exponentially, indicating the dominance of absorption, followed by a linear decrease signifying evaporation. In contrast, dense surface exhibited a linear decrease in the apparent contact angle, dominated by pure evaporation with no absorption. Additionally, the influence of gravity on the apparent contact angle of the droplet was investigated for different Bond numbers. The results indicated that an increase in the Bond number led to a decrease in the apparent contact angle. A modified Shapiro theory incorporating droplet height was developed to predict equilibrium contact angles for different Bond numbers. The present study explored the dynamics of apparent contact angles on downward-facing tubular and flat surfaces, focusing on droplet stability before detachment. A critical Bond number, found to be 3.5, characterized the event of droplet detachment. Finally, the study examined the impact of centrifugal forces on droplet contact angles for different fiber geometry and hydrophobicity. It was observed that larger droplets required lower rotational speeds for ejection, with hydrophobic surfaces facilitating easier droplet detachment. The findings of this study highlight the potential for controlling droplet behavior in various industrial and scientific applications, using dynamic effects
Non-unitary matrices embedding in quantum framework
No full textIn quantum computing, every operation must be unitary to ensure reversibility and preserve probability amplitudes, which is a fundamental requirement for any quantum process. However, many real-world computations in areas such as machine learning, signal processing, and dynamical modeling inherently involve non-unitary matrices. Since these cannot be directly implemented on quantum hardware, a practical workaround is required. One promising solution is block-encoding, a framework that embeds a non-unitary matrix A in C^{n x n} into a larger unitary matrix U_A in C^{2n x 2n}, enabling the simulation of non-unitary actions within quantum circuits through ancilla preparation and post-selection.
This thesis presents a novel and efficient method for block-encoding based on Gram-Schmidt orthogonalization. The algorithm constructs a unitary U_A with A embedded as the top-left block and remaining components filled to ensure global unitarity. Compared to traditional Singular Value Decomposition (SVD)-based approaches, the proposed method achieves at least 20% improvement in execution time, without compromising output fidelity or numerical stability. A corresponding quantum circuit is developed to extract A|psi> (that is, the transformed version of the input quantum state |psi> after applying the linear operator A) using post-selection on ancilla qubits, with the success probability determined byA|psi>^2.
To address the often low success rates inherent in post-selection schemes, the second phase of this research incorporates Grover-style amplitude amplification. Experiments on real quantum hardware demonstrate the integration of these techniques, supported by hardware-aware error mitigation strategies. Furthermore, the constructed unitary U_A is decomposed into native quantum gates for NISQ-era implementation.
The proposed block-encoding framework is applied to quantum machine learning, enabling the direct embedding of non-unitary matrices such as covariance, kernel, or weight matrices into quantum circuits. This allows efficient similarity estimation, matrix–vector multiplication, and other core operations, demonstrating its versatility in accelerating quantum linear algebra and supporting scalable, data-driven learning models on near-term quantum hardware