https://www.um.edu.mt/library/oar/handle/123456789/5626 Tue, 10 Nov 2026 07:11:12 GMT 2026-11-10T07:11:12Z https://www.um.edu.mt/library/oar/handle/123456789/150522 Title: Novel data analytics techniques for analysing satellite imagery data, traffic and environment pollution data to predict the hospital admission rates Abstract: Accurately forecasting hospital admissions is critical for optimizing healthcare resource allocation and emergency preparedness. This study presents a comprehensive, data-driven framework for predicting emergency department (ED) admissions related to respiratory, circulatory, and all-cause health conditions using local ground stations and satellite-based meteorological and air pollution data. Despite growing evidence linking environmental factors to hospital admissions, traditional hospital management systems lack integration with environmental intelligence and encounters problem in utilizing resources. To bridge this gap, our research employs advanced machine learning and deep learning models including tree-based, neural network, and regularized regression techniques to capture complex, nonlinear relationships between environmental exposure and health outcomes. A key contribution of this work is the introduction of three novel prediction schemes: Deep Explainable AI For Respiratory-Based Prediction (DXAIRP) for respiratory admissions, Deep Explainable AI For Circulatory-Based Prediction (DXAICP) for circulatory admissions, and Explainable Extreme Gradient Boosting Scheme for Environmental Impact on All-Cause Emergency Admissions (BGXAI) for all-cause admissions. These schemes were rigorously evaluated across multiple datasets, demonstrating high predictive performance and reliability. The analysis revealed that air pollution parameters (PM2.5, PM10, CO, NO2, and O3) are the most significant predictors of hospital admissions, while meteorological factors such as wind speed, temperature, and relative humidity also contribute, albeit to a lesser extent. The study further identified optimal lag times for different admission types, with a two-day lag providing the highest accuracy for respiratory and circulatory admissions, and a three-day lag optimizing all-cause admission predictions. Additionally, results indicate that ground-based environmental data yield superior predictive accuracy compared to satellite-derived measurements. To ensure interpretability and trust in AI-driven predictions, the framework integrates SHapley Additive exPlanations (SHAP), an Explainable AI (XAI) technique that elucidates the influence of each environmental factor on model outputs. The Extreme Gradient Boosting Regressor (XGBR) and Multi-Layer Perceptron Regressor (MLPR), and their ensembling with XAI establishes a robust, transparent, and adaptable predictive approach. Beyond its academic contributions, this research has significant practical implications. The predictive framework can be integrated into hospital management systems to support data-driven decision-making, enabling healthcare administrators to anticipate ED surges and optimize resource distribution. This framework includes ground station-based data to capture localized environmental conditions and also integrates satellite-based environmental data, making it particularly valuable for regions with limited ground monitoring infrastructure, thereby addressing spatial and temporal gaps in environmental health monitoring. Overall, this research advances climate-health analytics by offering a scalable, interpretable, and data-driven methodology for predicting hospital admissions based on environmental factors. The findings contribute to the growing field of AI-driven healthcare forecasting, with potential applications in national health systems, expansion to other disease categories, and deployment in resource-limited settings through cloud-based infrastructures. Description: Ph.D.(Melit.) Wed, 01 Jan 2025 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150522 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150027 Title: Determinants of acceptance into inpatient geriatric rehabilitation programmes in Malta : a regression analysis Authors: Muscat, Francesca; Ferry, Peter; Attard, Conrad; Lungaro-Mifsud, Stephen Abstract: With limited bed availability at Malta's state-run rehabilitation hospital, efficient and accurate assessment of frail older adults is essential for appropriate inpatient geriatric rehabilitation (GR) admissions. This study aimed to identify key predictors influencing admission decisions for inpatient GR programmes in Malta. Prospective data were collected from 250 patients aged 60 and over. In total, 83 predictors, including demographic details, clinical history, functional levels and social factors, were analysed against admission outcomes. Logistic regression analysis, mainly univariate analysis, followed by multivariate analysis with backward stepwise elimination, was used to determine significant predictors. Seven key predictors were identified, which shed light on the decision-making processes of the assessing specialists, highlighting the importance of embracing the ICF model rather than relying solely on medical criteria. These predictors reflect the WHO ICF and CGA frameworks, supporting a holistic, patient-centred approach. They informed the development of a digital tool, incorporating a future predictive model to guide GR admission decisions in Malta, aiming to facilitate the delivery of stratified and personalised care by enabling clinicians to select a care pathway based on the patient's potential health trajectory in response to rehabilitation interventions. Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150027 2026-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150026 Title: A turnaround time dashboard for hospital laboratory blood tests Authors: Camilleri, Sephora; Attard, Conrad Abstract: Turnaround time (TAT) is a critical performance indicator in laboratory medicine, directly influencing clinical decision-making and patient outcomes. Delays in reporting can compromise timely treatment, particularly for urgent or life-threatening conditions where rapid laboratory results are essential. Despite this importance, most laboratories at the local main hospital monitor TAT retrospectively and primarily through median values, offering limited insight into workflow inefficiencies. This study developed and evaluated a business intelligence dashboard for retrospective TAT monitoring with a conceptual extension to near real-time analysis, aligning with digital health initiatives that leverage data-driven tools to enhance healthcare quality and patient safety. Data extracted from the Laboratory Information System (LIS), comprising approximately 6 million test order records over 4 months, was transformed into key performance indicators, including median TAT, compliance with thresholds, and phase-specific data. Near real-time monitoring was defined conceptually, based on proposed frequent scheduled data refreshes rather than continuous live streaming. Semi-structured interviews with laboratory managers informed indicator selection and guided user-centred dashboard design. Usability was evaluated through structured testing with laboratory managers using the System Usability Scale (SUS), alongside task-based assessment of clarity, relevance, and ease of use. By transforming raw LIS data into actionable insights, the dashboard demonstrated the capacity to identify delay patterns and workflow inefficiencies that are not apparent in traditional retrospective reports, thereby supporting proactive operational decision-making. Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150026 2026-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/150020 Title: Simulation-based modelling of human-robot collaboration in long-term care : healthcare professionals’ perspectives on semi-humanoid and humanoid robots Authors: Attard, Conrad; Sultana, Roberta Abstract: Global population ageing and healthcare workforce shortages have led to increasing interest in robotic technologies within long-term care (LTC) environments. Notwithstanding growing evidence on robotic acceptance, limited research has used simulation-based modelling of human-robot interaction (HRI) scenarios with healthcare professionals particularly those working with older persons, to prospectively appraise collaborative workflows, escalation protocols, and organisational readiness. This qualitative, scenario-based simulation-informed conceptual modelling study, conducted as part of the HEARTS (Human-Centred Elderly Care: Semi Humanoid Robots for Collaborative Daily Support) project, used scenario-based human-in-the-loop simulation with ten health care professionals (HCPs) experienced in older persons' care. Eight structured HRI simulation scenarios; four video-based and four discussion-based, were designed using a semi-humanoid robot (Pepper). These HRI simulation scenarios encompassed practical operational failure conditions to stress-test human-robot collaboration (HRC) dynamics and handover protocols. Participants expressed a preference for a complementary human-robot model, in which a human carer works along with the robot instead of leaving it to operate independently. Results support the feasibility and analytical usefulness of simulation-based modelling as a methodological tool for pre-implementation workflow mapping. The study presents a state-based Human-Robot Collaboration Simulation Framework (HRCSF) identifying collaboration states, transition triggers and performance metrics for sustainable robotic integration in LTC environments. Thu, 01 Jan 2026 00:00:00 GMT https://www.um.edu.mt/library/oar/handle/123456789/150020 2026-01-01T00:00:00Z