Masters Theses

Energy drink consumption has become increasingly popular in Cambodia since theintroduction of Red Bull in 1997. High in caffeine and sugar, energy drinks are often marketed as energy boosters but are associated with various health risks, including cardiovascular diseases, neurological disorders, and metabolic issues such as diabetes. While existing studies have primarily focused on the health impacts of energy drinks in developed countries, there is a lack of qualitative research exploring the socio-cultural and economic factors influencing energy drink consumption in developing countries like Cambodia. This study is a qualitative study using in- depth interviews and observations to collect data. The result demonstrates that numerous factors shape their energy drink consumption behavior, including social norms, peer influence, economic factors, and accessibility. Simultaneously, some of them have attempted to reduce or stop energy drink consumption, aligning with the health action process, with emphasis on the transition from intention to action in health behavior. This finding innovatively bridges the gap between energy drink consumption and diabetes from polybasic aspects and provides solid evidence for policy making in the realm of global health in Cambodia

The post-COVID inflationary surge and subsequent disinflation have generated significant debate regarding their causes and policy implications. This paper assesses the drivers of inflation from the onset of the COVID-19 pandemic through the subsequent disinflation, with a particular focus on the sharp decline in inflation from June to July 2022. Using the Consumer Price Index (CPI), this analysis evaluates six primary explanations for inflationary trends: energy price shocks, supply chain disruptions, firm pricing power, wages and worker bargaining power, inflation expectations, and government stimulus.

A Rolling-Window Granger Causality model is employed to assess the temporal relationships between inflation and key economic indicators, allowing for a structural break in the data coinciding with initial stay-at-home orders. This approach identifies real personal expenditures, and the labor leverage ratio as significant contributors to the initial inflationary surge, while over-water shipping costs and the Federal Funds Rate exhibits a delayed relationship with disinflation. A time-interaction linear model is further applied to test the impact of identified economic events on inflation trends, revealing a lack of support for the common narrative that corporate profits, consumer income, or government transfers significantly drove inflation.

The findings suggest that inflation was largely driven by real economic shifts rather than speculative price-setting behavior, and that Federal Reserve policy may not have been the primary driver of disinflation. These results highlight the need for a more nuanced approach to assessing the causes of inflation and the efforts to mitigate it.

AbstractObjective To develop a 3D-printed phantom that accurately replicates patient-specific anatomical geometry, tissue textures, and attenuation characteristics derived from CT scans, thereby enabling precise lung tissue properties simulation for radiological medical imaging applications. Materials and Methods A streamlined workflow was developed to convert DICOM-format CT images into printer-executable G-code, eliminating conventional segmentation and intermediate file formats (e.g., STL). Using fused deposition modeling (FDM) with a 0.2-mm nozzle and polylactic acid (PLA) filament, the algorithm dynamically adjusts nozzle speed (5–50 mm/s) and extrusion rates to control line width (0.1–1 mm), thereby managing voxel density and replicating Hounsfield Unit (HU) values (-900 to 100 HU). Validation experiments used patient-specific lung CT data, and phantom accuracy was assessed through geometric measurements and HU value comparisons between printed models and original patient scans. Results The printed phantoms demonstrated a linear correlation (R² > 0.95) between designed fill rates and measured HU values, achieving submillimeter geometric accuracy in replicating lung vasculature and parenchymal structures. Manual measurements of 10 regions of interest (ROIs) revealed less than 5% deviation in dimensional fidelity, while HU distributions in phantom scans matched patient data within clinically acceptable margins (±50 HU). The method successfully simulated heterogeneous tissue textures, with printer parameters allowing for precise control over density gradients critical for radiometric applications. Conclusions This study introduces a breakthrough in patient-specific phantom fabrication, providing a rapid and cost-effective solution for validating CT-based techniques without exposure to ionizing radiation. The direct DICOM-to-G-code workflow ensures high anatomical and radiological fidelity and has applications in radiotherapy dosimetry, imaging protocol optimization, and medical training. Future research will expand to include multi-material printing and dynamic motion simulation to improve physiological realism.

Cardiovascular disease (CVD) remains the leading cause of mortality worldwide, with non-invasive imaging playing a crucial role in its diagnosis and management. However, optimizing these imaging technologies to improve patient outcomes remains a challenge. Virtual imaging trials (VITs) provide a powerful alternative to traditional clinical studies by using computational phantoms, virtual patients that can be imaged with simulated scanners, to systematically evaluate imaging technologies. The 4D extended Cardiac-Torso (XCAT) phantom series is one of the most widely used computational phantoms, providing anatomically detailed whole-body models with cardiac and respiratory motion. Despite their utility, the current XCAT cardiac models are limited in their ability to represent population variability and physiologically informed motion. The existing models are derived from gated 4D CT data of only two healthy individuals (one male, one female), restricting their ability to capture diverse anatomical and functional variations. Furthermore, abnormal cardiac motion is introduced manually, lacking a physiological basis and reducing realism and scalability. To address these limitations, this work explores techniques to develop a new series of 4D beating heart models derived from multiple sets of 4D cardiac-gated CT data. By segmenting and analyzing patient-specific cardiac motion, we first investigate an image-based method to construct a population of anatomically variable heart models that better reflect real-world variability, comparing the motion patterns to values reported in the literature. To further enhance realism, we explore a workflow for generating controlled variations in both normal and abnormal cardiac motion using finite element (FE) simulations. This approach integrates patient-specific electrophysiology and morphological parameters, allowing for the generation of physiologically informed cardiac motion models. These enhanced models can be integrated into whole-body XCAT phantoms to provide a novel population of virtual subjects with realistic anatomical and functional variability. The models can provide a valuable tool for virtual imaging trials, enabling the evaluation of emerging cardiac imaging technologies on a representative range of patient anatomies and motion patterns.

This study is motivated by the increasing need to mitigate storm-induced urban flooding, particularly in highly developed environments near floodplains or humid coastal regions, where climate change has intensified extreme rainfall events. The study explores a top-down approach to stormwater drainage system design, contrasting with the conventional bottom-up implementation commonly seen in urban planning.Two drainage pipe network layouts are evaluated within the Storm Water Management Model (SWMM) under identical storm conditions: one with a 4:1 inlet-to-outlet ratio per junction (shorter total flow path) and another with a 2:1 inlet-to-outlet ratio per junction (longer flow path with intermediate layer of pipes). Performance is compared based on runoff-outflow delay and internal system storage volume, using a controlled study area with consistent input parameters. To improve simulation accuracy, additional pipe flow networks replicating overland flow paths are incorporated, ensuring the full water movement process is represented—from surface accumulation to final discharge. Findings indicate that reducing in-system flow path length minimally impacts outflow delay, whereas shortening overland flow paths by increasing drainage inlets significantly improves performance. Results also suggest that dynamic storage volume does not correlate with discharge flow rate, indicating potential for increasing internal storage through intermediate pipe layers and extended conduit paths without negatively affecting system efficiency.

The clinical adoption of auto-segmentation tools has been hindered by the poor generalizability and interpretability of AI models, highlighting the need for automated contour quality assurance (QA) systems to verify the accuracy of auto-generated contours. A novel QA workflow has been proposed based on content-based image retrieval (CBIR), wherein the core idea is to retrieve image patches from a curated reference database that are visually similar to query patches sampled from a new scan, allowing the auto-contour segments in the query patches to be evaluated against the manual contours in the retrieved references.

To support this approach, we investigated the feasibility of quantifying visual similarity between 3D image patches using both handcrafted and deep radiomic features. Similarity prediction was formulated as a regression task, with the spatial distance between patch centers serving as a surrogate similarity label. A total of 90,000 patch pairs were sampled from 100 CT scans in the TotalSegmentator dataset. For each pair, 112 handcrafted radiomic features were extracted following the Image Biomarker Standardisation Initiative guidelines. These included 18 intensity-based features and 94 texture descriptors derived from GLCM, GLRLM, GLSZM, GLDZM, NGTDM, and NGLDM matrices. Deep radiomic features were extracted using a custom 3D convolutional neural network embedded in a Siamese architecture, trained end-to-end to predict spatial distances.

We first evaluated individual handcrafted features using univariate statistical analyses, including Spearman's rank correlation and divergence metrics, to assess their ability to distinguish visually similar versus dissimilar patches. Multivariate models, namely random forest and XGBoost regressors, were then trained on the full feature set. To identify the most informative features, we further applied selection techniques including Spearman correlation, permutation importance, Lasso regression, and principal component analysis. Among the regression models, the XGBoost regressor trained on handcrafted features achieved the best predictive performance, with a mean squared error (MSE) of 3.23 voxels and an r2 value of 0.800. In comparison, the Siamese network that extracted and exploited deep features achieved an MSE of 5.14 voxels and r2 of 0.693 after applying an outlier rejection strategy to improve prediction consistency under small spatial perturbations.

Our findings demonstrate that handcrafted radiomic features, when combined with machine learning models, offer an effective approach for modeling visual similarity in medical images. Although deep radiomic features currently yield lower performance, they remain promising due to their capacity for task-specific representation learning and the potential for further architectural refinement. Overall, this work supports the feasibility of developing a radiomics-driven CBIR framework for interpretable and localized contour QA in radiotherapy. Future work will focus on incorporating additional feature types and anatomical priors, refining similarity metrics beyond spatial distance, and validating the system on datasets with simulated or expert-annotated contouring errors to assess its clinical utility.

A dynamic treatment regime (DTR) is a sequence of decision rules, one per decision point, that map individual characteristics to a recommended treatment. An optimal DTR yields the largest mean utility if applied to select treatments in the target population. In this thesis, we consider estimation of an optimal DTR when the treatments consist of many smaller components, e.g., these components might be different aspects of a personalized message targeting some positive behavior change. We introduce a novel design for estimating optimal DTRs in randomized study, where each randomization stage includes a fractional factorial design. We introduce methods for estimating an optimal regime in both the single-stage and multi-stage setting. We derive the estimator for the optimal regime based on the potential outcome framework and Q-learning. We conduct simulation studies to evaluate the performance of the methods. A case study with single-stage design is used as an illustrative example.

Black choreography is integral and vital yet underrepresented undercurrent in the history of American modern dance. This essay provides the historical foundation of modern dance and Black people's positionality within the lineage by analyzing dance historian contributions. The essay then shifts to analyzing traditional archival research and shifts to de-center colonial and racist archival practices. The archival methodologies I describe offer a continuation of community and liberatory archival practices to preserve, innovate, and integrate into existing Black choreographic archiving. In response to my observations in the archive and North Carolinian presentations of Black women choreograph, I have created a choreographic and curated installation. I curated text, video, photography, and edited images with Solomon Thuo to apply the curatorial and archival methods I have researched. Through There is a ladder. and my archival instigation, I embody and demonstrate a choreographic liberatory practice that explores Black women perspective beyond archival limitations. This essay provides the introduction, positionality, historical contexts, and my thesis offerings to Black feminist choreography and archive practices in dance studies.

BackgroundCyanobacterial blooms pose a global environmental challenge, with Lake Taihu being one of the most severely affected regions. Microcystis aeruginosa is the dominant cyanobacterial species responsible for these blooms in Lake Taihu. Temperature is considered a key environmental driver influencing the formation of M. aeruginosa blooms; however, the molecular mechanisms underlying its temperature adaptation remain unclear. Previous studies have suggested that medium- and long-chain fatty acids (MLCFAs) play a crucial role in maintaining membrane stability and regulating energy metabolism, potentially contributing to temperature adaptation. To systematically investigate the adaptive mechanisms of M. aeruginosa under different temperature conditions, this study employed multi-omics approaches to integrate various biological data and elucidate regulatory processes at different levels. Methods In this study, we first measured the growth rates of M. aeruginosa at different temperatures and analyzed the composition of fatty acids under each condition. Pearson correlation analysis was performed to examine the relationship between growth rate and fatty acid content, and transcriptomic data were used to explore the gene expression patterns associated with fatty acid biosynthesis. Furthermore, weighted gene co-expression network analysis (WGCNA) was applied to identify key gene modules associated with temperature adaptation, followed by Gene Ontology (GO) enrichment analysis using TopGO to infer potential biological functions. Finally, an exogenous fatty acid supplementation experiment was conducted to validate the role of specific fatty acids in promoting M. aeruginosa growth under low-temperature conditions. Results Our results showed that the growth rate of M. aeruginosa began to accelerate at 25°C and reached its peak at 29°C. The composition of MLCFAs varied significantly across different temperatures, and correlation analysis revealed a strong positive relationship between total fatty acid content and growth rate. Further analysis of fatty acid metabolic pathways suggested that temperature changes led to increased fatty acid consumption, affecting their intracellular accumulation. The validation experiment confirmed that the supplementation of C18:3n3(9,12,15), C20:3n6(8,11,14), and C20:3n3(11,14,17) significantly enhanced the growth rate of M. aeruginosa at low temperatures (15°C). Transcriptomic analysis revealed extensive transcriptional reprogramming under different temperature conditions, particularly in pathways related to fatty acid biosynthesis, energy metabolism, and stress responses. WGCNA identified key gene modules strongly associated with fatty acid metabolism, suggesting potential regulatory mechanisms involved in temperature adaptation. Conclusion This study provides novel insights into the molecular basis of M. aeruginosa temperature adaptation by elucidating the interplay between growth, fatty acid metabolism, and gene expression regulation. The findings highlight the critical role of fatty acid metabolism in temperature adaptation and reveal coordinated regulation at the gene-transcript-metabolism level in response to temperature variations. These results contribute to a deeper understanding of cyanobacterial bloom dynamics and offer valuable knowledge for developing strategies to manage M. aeruginosa blooms in eutrophic water bodies.

Gypsum is a widely used sulfate mineral, and it is an important construction material used to make plaster and drywall due to its rapid hardening and low thermal conductivity. However, its mechanical performance under varying stress conditions, particularly the influence of water content on strength and deformation behavior remains underexplored. This study investigates the mechanical properties of gypsum through triaxial testing, focusing on stress-strain response, failure modes, and the Mohr-Coulomb failure criterion. Wet gypsum specimens with pore saturations of 10%, 20%, and 40%, were subjected to controlled loading conditions (0.414 MPa) to evaluate the impact of moisture on mechanical strength. Experimental results demonstrate that increased water content reduces peak strength. X-ray micro-computed tomography (Micro-CT) was used to analyze internal microstructural changes, revealing variations in particle distribution, and observing cementations formed inside of the gypsum samples. This study offers better insights into the mechanical response of gypsum, supporting its effective application in geomechanics and engineering practices.

While psychoanalysis seeks to understand the internal workings of the mind, law regulates external behavior, yet both disciplines shape and reflect human experience. This paper examines the transition from The Rule of Man to The Rule of Law from a psychoanalytic perspective, exploring how both traditional democracies and authoritarian states can achieve this transformation, while others remain trapped in cycles of failed democratization. This study argues that in stalled transitions, authoritarian leaders with narcissistic personality traits align with the fundamental characteristics of The Rule of Man, using large-scale political campaigns to inflict collective trauma. This trauma erodes the societal foundation necessary for legal and institutional transformation, preventing the establishment of The Rule of Law and leaving societies in a persistent liminal state. Through an analysis of Stalin’s totalitarian rule and Pinochet’s dictatorship, this paper examines the psychological mechanisms that reinforce authoritarian rule and hinder legal transitions. By integrating psychoanalytic theory with legal analysis, this research demonstrates how authoritarianism perpetuates itself through psychological manipulation, shaping governance and societal compliance, ultimately obstructing democratization efforts.

AbstractIntroduction: Adaptive Radiation Therapy (ART) enables real-time treatment modifications based on patient-specific anatomical changes, with online ART (oART) allowing daily plan adaptations using in-room imaging such as CBCT. The Varian Ethos system streamlines this process with automated contouring and plan re-optimization, necessitating robust secondary dose verification to ensure treatment plan calculation accuracy. Mobius3D and SciMoCa, two independent verification tools, employ different algorithms for dose calculation, with the 3%/2mm Gamma criterion commonly used for evaluation. This study aims to assess the accuracy of Mobius3D in oART by comparing it with SciMoCa and evaluating its sensitivity to intentional errors, ultimately refining quality assurance protocols for CBCT-based adaptive radiotherapy. Materials and Methods: 17 treatment plans, representing diverse anatomical sites and treatment techniques, were selected from Duke Health to ensure comprehensive validation. These plans were adapted from TrueBeam to Ethos in Eclipse Treatment Planning System (TPS) while maintaining field geometry and similar optimization objectives. Following re-optimization and quality assessment, the plans were exported to SciMoCa and Mobius3D for secondary dose calculations. Dose metrics (Dmean, D99%, D95%, and D1%) and gamma passing rates (3%/2mm) were evaluated. Additionally, phantom-based measurements were performed using Mobius Verification Phantom (MVP) and Delta4 to validate secondary calculations. Cases with large dose discrepancies between TPS and Mobius3D were further investigated by identifying the contributing factors in dose calculation accuracy, which included tissue heterogeneity and field off-centricity. To establish comparable plan, check criteria with other software and phantom measurements, different gamma criteria (1%/1mm, 2%/2mm, 3%/3mm, 5%/3mm) were also tested in Mobius3D and SciMoCa. To assess error sensitivity, treatment plans were modified by introducing isocenter shifts (2mm, 4mm) iv and collimator angle rotation (3°, 5°, 10°), followed by secondary dose verification using Mobius3D and SciMoCa. Statistical comparisons were performed to determine appropriate evaluation criteria and establish a consistent validation framework. Results: Mobius3D demonstrated an overall strong agreement with TPS calculations, achieving an average gamma passing rate of 91.7% at 3%/2mm across 17 cases. However, relatively large deviations were observed in breast cancer case, with the lowest passing rate of 70.1%, indicating potential limitations in breast cases calculations. Phantom-based measurements confirmed that both Mobius3D and Eclipse TPS slightly underestimated doses, with Mobius3D exhibiting larger deviations but remaining within clinically acceptable thresholds. Comparative analysis with Delta4 and SciMoCa revealed that Mobius3D consistently had lower gamma passing rates, with SciMoCa achieving an average of 98.6%, highlighting differences in algorithmic accuracy. Tissue heterogeneity and off-center field dose distribution were both identified as the contributing factors of dose discrepancy. Setting the entire body density to water equivalent improved Mobius3D’s gamma passing rates, confirming its sensitivity to inhomogeneous tissue effects. Additionally, iso-center adjustments to center the fields on the target in breast cases resulted in a significant increase in gamma passing rates, suggesting off-center fields as a contributing factor to lower dose calculation accuracy. Gamma criteria refinement showed that Mobius3D’s 5%/3mm criterion with a 90% threshold aligns most closely with SciMoCa and Delta4’s 3%/2mm standard, achieving a 94.12% passing rate, while the stricter 3%/2mm criterion resulted in only 64.71%. Dose metric analysis revealed that Mobius3D exhibited larger deviations in D99% (12 cases) and D95% (7 cases) compared to SciMoCa. The Dmean, with the smallest average difference (1.31%) and for 17 cases all within the 5% threshold compare with TPS, is the relatively stable matrix calculation in Mobius3D. Sensitivity testing showed that Mobius3D was highly responsive to geometric shifts, v with passing rates decreasing significantly for isocenter shifts (4mm reduced passing rate by 20%) and collimator rotations (10° shift reduced passing rate by 30.4%). The D99% and D95% values also decreased accordingly. These findings highlight Mobius3D’s ability to detect significant dose deviations but also its potential overestimation of minor errors compared to SciMoCa, reinforcing the need for refined acceptance criteria and optimization for adaptive radiotherapy applications. Conclusion: Mobius3D shows sufficient calculation accuracy overall but has limitations in inhomogeneous structures and off-center fields, where deviations are more pronounced. Dmean and the 5%/3mm 90% gamma criteria in Mobius 3D are more comparable with the results for SciMoCa and Delta4. Mobius3D is sensitive to error-introduced plans with iso-shift and collimator rotation, making it capable of detecting these deviations.

In this paper, I consider two arguments concerning the status of holobionts as evolutionary individuals—one rejects the thesis by privileging the “stability of lineages” (sensu Godfrey-Smith 2009) and the other supports the thesis by privileging the “stability of traits” (sensu Veigl et al. 2022). I argue that the tension between these two arguments arises from two fundamentally different accounts of natural selection. I suggest that each account of selection corresponds to a unique account of evolutionary individuality. This strategy entitles us to a modest pluralism: holobionts are evolutionary individuals on one account of selection but not on the other.

Purpose:Limited-angle CBCT reconstruction often suffers from incomplete projection data, resulting in severe wedge artifacts, image distortions, and reduced image quality. This study introduces a deep learning-based projection extrapolation filter to help high-quality CBCT image reconstruction from limited-angle data, aiming to mitigate artifacts and improve clinical usability.

Methods:This study developed a deep convolutional network based on the ResUNet architecture to extrapolate missing projection data. The training data are projections generated using TIGRE (Tomographic Iterative GPU-based Reconstruction Toolbox). The study simulates CBCT projections for 10 patients using TIGRE, generating projections over 180 degrees + fan angle (full-fan geometry) and 120 degrees (limited-angle geometry) to replicate real-world imaging conditions. Then the projections are resampled in angular dimension into a total of 7680 sinogram pairs (limited-angled and adequate-angled) that are randomly divided into training and validation sets in a 9:1 ratio with the remaining data reserved for testing. A ResUNet model is trained to extrapolate the limited-angled sinogram to adequate-angled sinogram. After the extrapolated data is resampled into projections, the final reconstruction was performed using the Feldkamp-Davis-Kress (FDK) algorithm. While focusing on reconstructed image quality and artifact reduction, performance metrics such as peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM) were used to quantify image quality improvements. Simultaneous attention to reconstruction image quality and artifact reduction.

Results:The proposed method can effectively generate the extrapolated projections with reduced image artifacts. The quantitative results showed the PSNR (33.012) and loss (0.002) of the model, which indicated a superior performance. The reconstructed CBCT volumes demonstrate superior image quality compared CBCT reconstructed with conventional methods using limited-angle data, and significantly reduces image artifacts. supporting the potential of integration in real-time clinical workflows.

Conclusion:Our deep learning-based projection extrapolation filter enables artifact reduction in CBCT reconstruction from limited-angle data. The proposed method holds promise for improving CBCT imaging quality in applications such as image-guided radiotherapy. Our future work includes using updated models to further improve extrapolated image quality and clinical evaluation of the proposed technique is warranted.

Compliant mechanisms utilize elastic bending to deform in a predictable manner, whileoffering advantages such as monolithic construction, high motion precision, and zero back- lash compared to traditional rigid-body mechanisms. These advantages offer a significant impact in the aerospace industry, where cost, weight, maintainability, and complexity are significant bounding variables. This thesis presents the design and analysis of a bi-stable compliant mechanism utilizing beam bracing to achieve beam snap-through while a critical strain is exceeded. Parametric equations for the response of this mechanism are derived using pseudo-rigid body models (PRBMs) and potential energy to enable efficient iterations on initial designs and to inform parameter tuning for desired behaviors. These parametric equations provide an efficient alternative to more computationally expensive finite element methods, especially with regard to initial mechanism design. A potential application of the proposed compliant mechanism is in aerospace structural health monitoring (SHM). Integrating the bi-stable mechanism response as a function of specified strain threshold offers both a passive and active indicator of excessive loading or possible damage progression in principal structural elements (PSE). Experimental valida- tion and numerical simulations in ANSYS demonstrate the feasibility of the mechanism as an SHM sensor. This study contributes to the broader field of compliant mechanisms by providing a design and analysis methodology for a bracing-triggered bi-stable system that is activated by strain exceedance. This research also highlights the potential applications for compliant mechanisms in aerospace structural monitoring in both passive and active systems as a possible complementary addition to conventional sensors.

Introduction: The widespread adoption of Electronic Health Records (EHR) has revolutionized healthcare delivery while introducing critical challenges in health data security and privacy. Despite advancements, legacy EHR systems remain vulnerable to breaches and struggle to comply with stringent regulations like GDPR. Current blockchain-based solutions, while promising, suffer from non-minimalist integration, ambiguous ownership models, and hybrid storage vulnerabilities. This research addresses these gaps by proposing Privacy PIN, a decentralized, privacy-preserving plugin designed to retrofit legacy EHR systems for GDPR compliance. Privacy PIN integrates cryptographic sovereignty, biometric authentication, and blockchain-anchored data governance to empower patient-centric control over health data, bridging the divide between regulatory mandates and scalable digitization.Methods: This research simulates Privacy PIN on the Sepolia Ethereum Testnet sandbox to emulate Ethereum Mainnet. Black-box testing validates Privacy PIN output reliability under GDPR requirements. Performance analysis extends to capacity expansion solutions, including Polygon and BASE, measuring gas costs, block throughput, and cost-effectiveness to assess compatibility for massive adoption for global health. Delphi method is also applied to explore certain GDPR rights. Results: Privacy PIN has the ability to individually implement the eight object rights in the GDPR, right to be informed, right of access, right to data portability, right to rectification, right to be forgotten, right to restrict processing, right to object to processing and right in relation to automated decision making and profiling. Additionally, this research also proposes a technical framework for integration with existing EHR systems, accompanied by an example provided as a demonstrative case study in the appendix. Discussion: There are also limitations in the research: (1) ethical risks of health data NFTization enabling black market exploitation; (2) infrastructure gaps (e.g., unstable energy or networks in Sub-Saharan Africa) deepening healthcare disparities; and (3) rapid blockchain evolution like Solana jeopardizing Ethereum-based compatibility. Additional challenges include informed consent barriers for non-technical users and ethical oversight complexities. Conclusion: Privacy PIN provides a cost-effective solution for GDPR-compliant EHR modernization, leveraging decentralized identity, biometric authentication, and NFT-based data sovereignty. While constrained by infrastructural, ethical, and technological dependencies, its modular design enables incremental adoption across heterogeneous healthcare systems. Future work should prioritize the standardization of Web3 health identifiers, patient education initiatives, and policy frameworks to harmonize privacy preservation.

Background: Effective chronic disease management for elderly patients with hypertension anddiabetes remains a significant challenge in community health settings. Peer education has been widely promoted as a strategy to enhance patient engagement, but its feasibility and effectiveness in specific local contexts require further examination. This study explores the development and iterative refinement of a community-based health management strategy tailored to elderly patients in Kunshan, China. Methods: A qualitative study was conducted using semi-structured interviews with elderly patients, their family members, community health workers, family physicians, and health officials. Data were analyzed using a framework informed by the Health Belief Model (HBM), focusing on perceived barriers, self-efficacy, and cues to action in chronic disease management. Results: Findings revealed significant barriers to health behavior change among elderly patients, including physical limitations, economic burdens, and competing household responsibilities. The study identified critical challenges that hindered the successful implementation of peer education strategies, such as low health literacy, limited peer leadership capacity, and preferences for professional guidance. In response, participants expressed receptiveness to technology-assisted health education tools. Voice-bot interventions were proposed as a feasible solution to enhance patient engagement and reduce healthcare provider workload. Conclusion: This study highlights the limitations of peer education models in the context of Kunshan and underscores the need for adaptable, technology-supported community health interventions. Future research should focus on patient-centered evaluations of Voice-bot interventions to assess their effectiveness in improving chronic disease management among elderly patients.

Purpose: To develop a computational framework for patient-specific dose agreementverification in cone-beam CT (CBCT) -guided radiation therapy (RT) for esophageal cancer treatment. Materials/Methods: Ten esophageal cancer patients undergoing CBCT-guided RT (Total Dose = 50 - 60 Gy, Treatment Fraction(Fx) =25 - 30, 5 Fx/ week) were retrospectively studied, and treatment plan, planning CT (pCT), CBCT images prior to each treatment fraction, along with couch correction records were collected for each patient. Two computational experiments were designed to calculate (1) fractional dose without CBCT guidance, and (2) fractional dose distributions under CBCT-guided RT. In the first experiment, a series of virtual CT were generated by rigidly registering the pCT to each CBCT images, mimicking patient anatomy after initial patient setup through laser system and surface marker. The original treatment plan was then applied to compute the fractional dose distribution. In the second experiment, couch movement data, as guided by CBCT, were further incorporated to refine the final patient positioning and fractional dose distributions were calculated similarly. The total dose can be calculated by combining fractional doses with/without CBCT for every patient, simulating different clinical scenarios such as full CBCT utilization, missing 5, 10, 15, 20 CBCT throughout the treatment, and complete absence of CBCT, as in traditional methods different from image-guided radiation therapy (IGRT). For each scenario, 100-fold random combinations were performed, and calculated total dose was compared with the planned dose based on the following clinical evaluation metrices: PTV ´ D2%, PTV ´ D98%, Lung ´ Dmean, SpinalCord ´ Dmax, and Heart ´ V30. Results: Our framework successfully quantified the impact of CBCT utilization on patient-specific dosimetric outcome. PTV coverage decreased for every 5 missing fractions of CBCT utilization. Dose to organ at risk changes depending on the relative position of target area with respect to the organ at risk. Conclusion: The proposed method provides a computational framework to verify patientiv specific dose agreement in CBCT-guided esophageal cancer RT. Our methods can also be generalized to other IGRT modalities across various treatment sites.

AbstractPurpose: Cone-beam computed tomography (CBCT) plays a critical role in radiotherapy by providing image guidance for accurate tumor localization. Current CBCT reconstruction methods rely on filtering techniques that introduce noise and artifacts. This study aims to evaluate the effectiveness of a deep learning model, ResUNet as an end-to-end solution in CBCT reconstruction, improving image quality and tumor localization accuracy. Methods and Materials: This study utilized the TIGRE toolbox to simulate CBCT forward projections from a high-quality CT volume image (GroundTruth-CT). The CBCT projections were reconstructed into a CBCT volume image (Raw-CBCT) using backprojection without filtering, resulting in a blurred and low-quality image. A supervised convolutional neural network (CNN) model, ResUNet, was trained to enhance Raw-CBCT by learning to predict the high-quality GroundTruth-CT. The input to the network was Raw-CBCT, and the ground truth was GroundTruth-CT. The model was optimized using the mean squared error (MSE) loss function. The final output of the model was an enhanced CBCT volume image (ResUNet-CBCT). For comparison, the standard Feldkamp, Davis, and Kress (FDK) reconstruction method was applied to the CBCT projections, producing FDK-CBCT. Both EN-CBCT and FDK-CBCT were compared to GroundTruth-CT using structural similarity index measure (SSIM), peak signal-to noise ratio (PSNR), and mean squared error (MSE) to evaluate image quality and reconstruction accuracy (Sara, Akter, & Uddin, 2019). Results: The ResUNet model significantly outperformed the standard FDK reconstruction method. For test patients (L333, L096), ResUNet achieved notably lower Mean Squared Error (MSE), higher Structural Similarity Index Measure (SSIM), and higher Peak Signal-to-Noise Ratio (PSNR), indicating improved image quality. Specifically, ResUNet demonstrated superior artifact suppression, reduced noise, and enhanced anatomical detail visibility compared to FDK reconstruction, especially in low-contrast regions. Conclusions: This study demonstrates the ResUNet model can effectively perform end to end CBCT reconstruction including replacement of the filtering step, resulting in superior image quality compared to standard FDK_CBCT. It will also lead to reduced imaging dose and efficiency. Future research will focus on optimizing network architectures and validating performance using larger clinical datasets to further advance CBCT imaging in radiotherapy.

The migration of healthcare workers, particularly the nurses, from low- and middle-income countries like Sri Lanka has emerged as a critical issue. It is aggravating workforce shortages and straining healthcare systems. This study explores the factors influencing Sri Lankan nurses’ decisions to either remain in the country or to migrate abroad. The aim of this study is to understand the push and pull factors that shape these decisions, focusing on the common critical period of indecision among new nursing graduates.

This study adapts a mixed-methods approach, combining surveys and in-depth interviews. Two separate batches of participants were involved in this study: the surveys were collected from 25 Sri Lankan nurse graduates who graduated during the years of 2010 to 2014, and in-depth interviews were conducted with 20 recent graduates who graduated during the years of 2021 to 2023 that are currently working in Sri Lanka. This study uses thematic analysis to identify key themes related to the nurses' migration decisions.

The findings reveal that while financial incentives and better working conditions abroad are strong pull factors, personal and familial responsibilities, and a sense of duty to the country, are significant factors in retention. Nurses express a desire to remain in Sri Lanka but are hindered by heavy workloads, low salaries, and limited career advancement opportunities. The results also emphasize that the government of Sri Lanka has a window of opportunity to intervene during a period of indecision by addressing these systemic challenges through policy changes.

Policy suggestions based on the findings of the study include increasing nurses’ salaries, improving career development pathways, and reducing workloads. The study also calls for greater international collaboration to regulate and monitor the global recruitment of healthcare workers to prevent further brain drain. The research contributes to a deeper understanding of the dynamics of nurse migration and retention, offering insights for policymakers to retain the Sri Lankan skilled workforce and to strengthen Sri Lanka’s healthcare system.