Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
5 results
Search Results
Item Restricted GUIDELINE-DIRECTED LIPID-LOWERING THERAPY IN U.S. ADULTS: TRENDS IN UTILIZATION, TREATMENT PATTERNS, DISEASE BURDEN, AND HEALTHCARE EXPENDITURES(Saudi Digital Library, 2026) Alharbi, Abdulrahman; Sherbeny, FatimahBackground. Lipid-lowering therapy (LLT) is among the most widely prescribed pharmacologic interventions in the United States and plays a central role in the prevention and management of atherosclerotic cardiovascular disease. The 2018–2023 period spans the maturation of the 2018 ACC/AHA cholesterol guideline, continued generic market expansion, and the COVID-19 pandemic, yet contemporary nationally representative evidence on utilization, treatment intensity, disease burden, and expenditures among LLT users remains limited. Objective. To characterize trends in LLT utilization, statin intensity, comorbidity burden, payer-stratified expenditures, and predictors of high-intensity statin prescribing among U.S. adults aged 40 years and older who used LLT. Methods. A repeated cross-sectional analysis of the Medical Expenditure Panel Survey (MEPS) 2018–2023 was conducted, restricted to adults aged ≥40 years reporting at least one LLT fill (unweighted N = 25,662; weighted ≈ 281.5 million person-years observations). Survey-weighted estimates were generated in Stata 19 using svy procedures. Trends were tested with survey-weighted regression; predictors of high-intensity statin use were modeled with survey-weighted logistic regression, and expenditures with generalized linear models (gamma family, log link). Results. Atorvastatin use rose from 51.2% to 56.9% of LLT users and rosuvastatin nearly doubled from 11.5% to 22.8% (both p < .001), whereas simvastatin (23.2% to 12.7%), pravastatin (12.0% to 6.2%), and lovastatin (4.5% to 2.3%) declined (all p < .001). Ezetimibe use increased modestly from 3.4% to 5.7% but remained low. High-intensity statin therapy rose from 31.9% to 40.8% (p < .001). Total national LLT expenditures fell 71.4%, from $17.2 billion to $4.92 billion, with private-payer spending declining most steeply (−82.1%) and Medicare remaining the largest single payer. Higher-intensity prescribing was independently associated with cardiovascular events (adjusted odds ratio [aOR] = 2.1), higher Charlson comorbidity burden (CCI ≥3: aOR = 2.04), public insurance (aOR = 1.16), and later survey year (2023 vs. 2018: aOR = 1.54), and inversely associated with female sex (aOR = 0.62) and age ≥75 years (aOR = 0.76). Conclusions. Between 2018 and 2023, LLT prescribing shifted toward higher-potency agents and higher-intensity statin therapy while aggregate spending declined sharply, driven largely by generic substitution. Persistent underutilization of ezetimibe and demographic disparities in high-intensity prescribing identify clear targets for guideline-directed quality improvement.15 0Item Restricted AI-Enabled Autonomous Knowledge Extraction from Large-Scale Textual Data(Saudi Digital Library, 2026) Alharbi, Abdulrahman; Obradovic, ZoranThe rapid growth of large-scale textual data across social media platforms, news media, and scientific repositories presents both unprecedented opportunities and significant challenges for extracting meaningful insights. During global events such as the COVID-19 pandemic, understanding public discourse requires analyzing vast amounts of noisy, heterogeneous, dynamic, and geographically distributed data. At the same time, the exponential increase in scientific publications has made traditional evidence synthesis methods increasingly labor-intensive, time-consuming, and difficult to scale. Existing approaches to textual knowledge extraction often operate in isolation, lack interpretability, fail to integrate heterogeneous data sources, and do not support scalable end-to-end automation. This dissertation addresses these limitations by proposing a unified framework for AI-enabled autonomous knowledge extraction from large-scale textual data. The research introduces a comprehensive pipeline that integrate sentiment analysis, topic modeling, semantic interpretation, spatiotemporal reasoning, and multi-agent automation for scalable, robust and reproducible text analysis across heterogeneous domains. First, this work introduces TriLex, a novel unsupervised sentiment analysis framework that combines multiple lexicon-based sentiment analysis methods through weighted aggregation, majority voting, and dynamic thresholding technique to improve robustness and accuracy for short and noisy textual data. Building on this foundation, a hierarchical spatiotemporal framework is developed to capture the evolution of public sentiment across global, national, and regional scales. The framework integrates over 7 million social media posts and thousands of news articles to analyze COVID-19 vaccine discourse across time, geographic regions, and platforms. To enhance topic interpretability, this research integrates BERTopic with large language models (LLMs), enabling automated generation of coherent and context-aware topic representations for large-scale textual discourse. A cross-platform analytical framework is further introduced to examine temporal relationships between social media and news media discourse, demonstrating a bidirectional relationship in which news coverage and public discourse influence each other over time. Extending beyond discourse analysis, this dissertation introduces an Agentic AI framework that automates the end-to-end process of large-scale multilingual knowledge extraction and evidence synthesis. The proposed multi-agent system coordinates specialized agents for query generation, multilingual retrieval, metadata harmonization, title and abstract screening, and full-text analysis. Evaluated on a multilingual corpus of over 52,000 scientific records, the framework achieves high screening performance while substantially reducing processing time from months to hours, demonstrating significant improvements in scalability, robustness, and reproducibility. Collectively, this dissertation bridges the gap between analytical understanding and autonomous knowledge extraction from large-scale textual data. By integrating robust sentiment analysis, interpretable topic modeling, spatiotemporal discourse analysis, and autonomous AI systems within a unified framework, this work establishes a scalable and extensible paradigm for transforming heterogeneous textual data into actionable knowledge. The proposed methodologies are validated using real-world datasets spanning social media, news media, and scientific literature across diverse application domains.13 0Item Embargo Multi-Model Deep Learning Approach for Sentiment Analysis of Saudi Dialectal Arabic(Saudi Digital Library, 2026) Alharbi, Abdulrahman; Sharma, NabinIn today’s digital era, social media platforms have become integral to modern life, providing corporations, governments, and decision-makers with valuable insights derived from large volumes of user-generated content. However, extracting meaningful information from unstructured textual data remains a significant challenge. To address this issue, text classification techniques, particularly sentiment analysis, have been widely adopted to evaluate public opinions and sentiment polarity. While extensive research has focused on English-language data, low-resource languages such as Arabic, characterised by complex morphology and rich dialectal diversity, have received comparatively limited attention. Arabic sentiment analysis is further complicated by linguistic ambiguity and regional variation, which reduce the effectiveness of conventional NLP methods, especially for informal social media content. Moreover, there is a noticeable lack of annotated datasets and comprehensive studies focusing on Arabic dialect sentiment analysis, with the Saudi dialect being particularly underrepresented. This thesis addresses these challenges by focusing on sentiment analysis of Saudi dialect social media content using a range of machine learning (ML) and deep learning (DL) techniques. The research begins with the collection and manual annotation of a dataset related to Saudi education reform from X (formerly Twitter). This is followed by a systematic evaluation of classical ML models using a wide range of feature extraction methods and pre-trained Arabic word embeddings, establishing strong baselines and identifying effective configurations for Arabic sentiment classification. To capture richer semantic and morphological information, the thesis then proposes a hybrid word embedding strategy that integrates pre-trained AraVec and FastText representations at both tweet and word levels, enabling the development of multiple DL architectures for improved sentiment classification. Experimental results demonstrate that different DL models capture complementary sentiment-bearing linguistic patterns. Motivated by this observation, the thesis introduces the CBiR-LR stacked ensemble, which integrates CNN, Bi-LSTM, and RNN base learners with a Logistic Regression meta-classifier to effectively combine their predictions, enhancing robustness and mitigating individual model biases. To complement the ensemble-based CBiR-LR approach, which exploits architectural diversity over static word embeddings, the thesis further investigates a context-aware modelling paradigm. Leveraging the availability of transformer-based Arabic language models, MAR-BiCAtt is proposed to explore the impact of contextualised embeddings and attention mechanisms on dialectal Arabic sentiment analysis. MAR-BiCAtt integrates MARBERT-generated contextual representations with Bi-LSTM, CNN, and attention components to model token-level semantic interactions and selectively focus on sentiment-relevant expressions. This design enables a principled comparison between ensemble learning over static representations and deep contextual modelling, providing insights into their respective performance characteristics and generalisation behaviour across multiple Arabic sentiment datasets. To evaluate robustness and generalisation, additional experiments were conducted on three public Arabic sentiment datasets spanning different domains. The results demonstrate that the proposed approaches achieve consistent performance and improved generalisation across diverse application areas, including education, healthcare, and socio-political sentiment analysis.21 0Item Restricted Delivering Prehabilitation in Cancer Surgery: A Service Evaluation at Nottingham University Hospitals(Saudi Digital Library, 2025) Alharbi, Abdulrahman; O’Connor, DominicAbstract Background: Cancer surgery carries high risks, with complications linked to delayed recovery and poorer outcomes. Prehabilitation aims to optimise patient fitness before surgery, yet evidence from real-world NHS settings remain limited. In 2022, Nottingham University Hospitals NHS Trust introduced a multimodal prehabilitation service, developed in line with Macmillan Cancer Support guidance. Methods: This service evaluation included 1,720 patients triaged to Specialised (n = 329), Targeted (n = 943), or Universal (n = 448) prehabilitation pathways. Outcomes assessed pre- and post-programme included functional capacity (incremental shuttle walk test [ISWT], 60-second sit-to-stand [STS] test, grip strength), psychological health (GAD-7, PHQ-9), and physical activity. Analyses used paired t-tests, ANOVA, and effect size calculations. Results: Significant improvements were observed across outcomes. ISWT increased by 57 m (p < 0.001, d = 0.6), STS by 6 repetitions (p < 0.001, d = 0.9), and grip strength modestly (p < 0.001). Anxiety (Δ –1.9) and depression (Δ –2.0) scores decreased (both p < 0.001, d ≈ –0.5). Weekly physical activity more than doubled (+142 min/week), and strength sessions increased by 2.4 sessions (both p < 0.001, d > 1.0). Between-group differences were limited, although PHQ-9 scores improved more in the Specialised than the Targeted pathway and strength sessions more in the Universal than the Targeted pathway. Conclusion: A multimodal prehabilitation programme delivered within routine cancer care was associated with meaningful functional, psychological, and behavioural gains. However, barriers to engagement highlight the need for flexible delivery models and systematic follow-up to maximise accessibility and sustainability.12 0Item Restricted DATA ANALYTICS FRAMEWORK FOR IMPROVING THE SAFETY AND CAPACITY OF AIRSPACE(Cranfield University, 2024-03-21) Alharbi, Abdulrahman; Petrunin, IvanDue to their flexibility and general robustness, unmanned aerial vehicles (UAVs), have increasingly been deployed for diverse applications. These include aerial mapping, surveillance, package delivery, and even agriculture. Increased employment, however, has also entailed new demands for smart, nimble and effective UAV traffic-management systems, particularly in urban areas. If numerous, fully automized UAVs are to be flown frequently, and beyond the visual line of sight (BVLoS), then efficient unmanned traffic management (UTM) is essential, not least as UAV traffic will inevitably become denser. In future, indeed, air-traffic management will also be more complex, and airspace more crowded, as the sheer volume of UAVs continues to rise. Consequently, UTM will require swift, efficient decision-making mechanisms. Important challenges also remain in terms of machine-learning algorithm verification, these stemming primarily from a lack of explicability and transparency. Given that traditional safety mechanisms are unequal to the tasks involved, this has been an inhibiting factor in the integration of UAVs into very low-level (VLL) airspace. This thesis aims to develop a data-analytics framework to characterize traffic-flow patterns of UTM airspace by analyzing simulated historical data. The pertinent data analysis supports risk analysis, and it also improves trajectory planning in different airspace regions. It considers all dynamic parameters, such as extreme weather, emergency services, and dynamic airspace structures. Furthermore, and to meet the critical need for accurate congestion prediction in UAS traffic flow management (UTFM), this study uses state-of-the-art machine learning techniques to integrate air traffic-flow prediction with the intrinsic complexity metric. In this study, air-traffic congestion analysis and prediction will be addressed via a deep-learning methodology, within a UTM context, across a timeframe of three minutes. The proposed model is distinct from approaches that would focus on the more conventional issues of conflict detection, conflict resolution and trajectory prediction. In addition, this thesis proposes a tailored solution to the needs of demand-and capacity-management (DCM) services. This solution deploys a transparency based methodology, with a fusion of both black-box and explainable, white-box models. It generates, therefore, an intelligent system that can be both explicable and reasonably comprehensible. The results show that the advisory system will be able to indicate the most appropriate regions for UAV operations, while increasing UTM airspace availability by more than 23%. Keywords:23 0