324139 research outputs found
Sort by
Application of formal modelling to mitigate metadata driven privacy violation
Get PDFIn this current Data Economy, the increasing value of data, in addition to the lack of ownership afforded to users, has brought about a plethora of digital privacy issues. These issues can be seen as a result of the rate at which data generated by users absconds to exist in the databases of data aggregators and brokers, often occurring without user awareness, conscious input, or reasoned consent. As it relates to passive data, which is mostly in the form of metadata, such data collection risks exposing users to inferential predictions made by sophisticated data algorithms on these aggregated data sets that reside in companies’ databases. n this dissertation, we have dissected the various challenging and, at times, conflicting issues in this area leading us to draft consequent subsidiary research questions to address these issues. From this, in this dissertation we consider the following question: “How can we leverage a decentralised approach to data ownership, together with formal models of data. access, to aid in the protection of data subjects’ privacy so as to mitigate inference-driven identity exposures from metadata collection?”Thus, our contributions are as follows. First, we explicitly outline our understanding of privacy. We then explore a formal model for the Solid ecosystem, which serves to provide a decentralised web architecture offering user-determined access control. Whereupon, we decided to build upon this decentralisation structure via the dynamic Category-Based Access Control (CBAC) framework.Finally, we present a formal model that has the potential to underpin models to mitigate inference-driven identity exposures from metadata collection through compounding the prior research contributions in a cohesive manner that meets our intended privacy objectives
Induced subgraph density. VII. The five-vertex path
No full textWe prove the Erdős-Hajnal conjecture for the five-vertex path ; that is, there exists such that every -vertex graph with no induced has a clique or stable set of size at least . This completes the verification of the Erdős-Hajnal conjecture for all five-vertex graphs. Our methods combine probabilistic and structural ideas with the iterative sparsification framework introduced in the third and fourth papers in the series
Large sieve inequalities for exceptional Maass forms and the greatest prime factor of
Get PDFWe prove new large sieve inequalities for the Fourier coefficients of exceptional Maass forms of a given level, weighted by sequences with sparse Fourier transforms – including two key types of sequences that arise in the dispersion method. These give the first savings in the exceptional spectrum for the critical case of sequences as long as the level, and lead to improved bounds for various multilinear forms of Kloosterman sums. As an application, we show that the greatest prime factor of is infinitely often greater than , improving Merikoski’s previous threshold of . We also announce applications to the exponents of distribution of primes and smooth numbers in arithmetic progressions
Forty years later: adult health and non-communicable disease following the 1984–1985 Great Ethiopian Famine – a retrospective cohort study
Get PDFBackground: As the threat of child malnutrition increases, the focus remains mostly on short-term consequences. Long-term sequelae are increasingly recognised but lack strong evidence, and many studies face methodological limitations. Method: A retrospective cohort of survivors of the 1984–1985 Great-Ethiopian Famine was compared with two novel control groups: born post-famine; and age category- and sex-matched controls. Exposure to famine at different age categories was explored (fetal, 0–2, 2–5, 5–10 and 10–18 years). Follow-up was 40 years later. Outcomes included anthropometry, body composition, arterial stiffness, mental health, and risk of cardiometabolic and non-communicable diseases (NCDs). Adjusted differences and 95% CI between exposed and controls were calculated. Results: Compared with matched and post-famine controls, adjusted differences (95% CI) for exposed group were: height, −1.4 cm (−2.4 to –0.3) and −2.4 cm (−3.7 to –1.1); weight, −1.4 kg (−2.7 to –0.1) and −1.7 kg (−3.3 to –0.1); diastolic blood pressure (DBP), −2.8 mm Hg (−4.4 to –1.1) and 2.8 mmHg (0.9 to 4.7); handgrip strength, −1.7 kg (−2.7 to –0.6) and −4.1 kg (−5.5 to –2.7); brachial augmentation index, 5.4% (0.3% to 10.5%) and 16.1% (10.1% to 22.1%); aortic augmentation index, 6.0% (1.5% to 10.4%) and 11.7% (6.1% to 17.3%); subscapular skinfold thickness, 1.1 mm (0.2 to 1.9) and 1.2 mm (0.1 to 2.3); triceps skinfold thickness, 1.8 mm (0.8 to 2.7) and 2.1 mm (1.0 to 3.3) and waist-to-height ratio, 0.01 (0.003 to 0.02) and 0.01 (0.001 to 0.02), respectively. When comparing risk by timing of exposure, individuals exposed during early childhood (0–2 years), preschool age (2–5 years), and late childhood (5–10 years) had reduced adult stature of −2.8 cm (–4.8 to –0.9), −2.8 cm (−4.7 to -0.9) and −2.1 cm (–4.0 to −0.2), respectively, and increased triceps skinfold of 1.7 mm (−0.5 to 3.8), 3.2 mm (0.8 to 5.6) and 3.8 mm (1.6 to 6.02), respectively. Conclusions: Early-life famine exposure is associated with smaller adult size and several, but not all NCD risks. Lower DBP in survivors compared with matched controls is surprising and might reflect differential susceptibility to specific later-life health risks. Greater arterial stiffness underscores the need to identify both preclinical and clinical risk. In contrast to exposure in utero, risk was higher among those exposed during early childhood (0-2 years), preschool (2–5 years) and late childhood (5–10 years). The study underscores the need for a dual approach in low- and middle-income settings: tackling the immediate undernutrition while also anticipating and mitigating long-term NCD risk in populations exposed to early-life severe malnutrition or famine
Comparing Efficacy and Safety of Bridging Therapy Versus Endovascular Thrombectomy in Acute Basilar Artery Occlusion: A Systematic Review and Meta‐Analysis
Get PDFBackground: Basilar artery occlusion (BAO), a rare and severe stroke, causes high morbidity and mortality. This meta‐analysis aims to compare bridging therapy, including endovascular thrombectomy (EVT) with intravenous thrombolysis (IVT), versus EVT alone in BAO. Methods: PubMed, Cochrane Central, and ScienceDirect were searched until May 2025. The risk ratios (RRs) and 95% confidence intervals (CIs) were combined using a random effects model in Review Manager software. The quality assessment was conducted using the Cochrane Risk of Bias (RoB 2.0) and the Newcastle–Ottawa scale. Publication bias was assessed visually via funnel plots and statistically using Egger's regression test. This review's protocol was registered on PROSPERO with the ID: CRD420251108752. Results: A total of 14 studies, including 11 observational studies and 3 randomized controlled trials, with 3745 participants, were analyzed. Bridging therapy was associated with a significantly higher likelihood of achieving functional independence (modified Rankin Scale (mRS) ≤ 2) (RR = 1.27; 95%CI: [1.13, 1.43]; p < 0.0001; I2 = 23%) and independent ambulation (mRS ≤ 3) (RR = 1.10; 95%CI: [1.01, 1.20]; p = 0.02; I2 = 0%). The mortality risk was also significantly lower in bridging therapy (RR = 0.83; 95%CI: [0.75, 0.93]; p = 0.001; I2 = 0%). The successful recanalization endpoint was found to be comparable between the bridging therapy and EVT alone arms (RR = 1.00; 95%CI: [0.97, 1.03]; p = 0.99; I2 = 0%). Similarly, there was no significant difference between the intervention and control groups in spontaneous intracranial hemorrhage (sICH) (RR = 0.99; 95%CI: [0.71, 1.39]; p = 0.97; I2 = 0%). Conclusion: Our meta‐analysis supports administering IVT before EVT for BAO, showing benefits in functional outcomes and mortality without increasing hemorrhage risk, although successful recanalization was similar in both groups
Cerebrolysin after Endovascular Thrombectomy in Stroke: 12‑Month Functional Outcomes in a Propensity‑Matched Cohort
Get PDFBackground Long-term outcomes after endovascular thrombectomy (EVT) and the role of adjunctive neuroprotection to achieve post-stroke independence remain incompletely characterized. In this hypothesis-generating target-trial emulation, we assessed 12-month functional outcomes in a prespecified extension of a propensity score–matched cohort of rigorously selected EVT patients treated with adjunctive Cerebrolysin, a multimodal neuroprotective agent. Methods Consecutive EVT patients were prospectively enrolled and treated with Cerebrolysin 30 mL/day for 21 days starting immediately post-EVT, with a second 21-day course at 69–90 days. Outcomes were compared with historical controls using 1:1 nearest-neighbor propensity score matching on ten prespecified covariates. The primary endpoint was functional independence (modified Rankin Scale [mRS] 0–2) at 12 months. Secondary endpoints included 12-month mRS shift, Barthel Index (BI), and need for institutional care. Multivariable regression models were used to estimate adjusted associations, with prespecified sensitivity analyses including calendar time and key EVT predictors. Results Cerebrolysin use was associated with higher odds of 12-month functional independence after adjustment for potential confounders (aOR 6.10, 95% CI 1.64–22.66; p<0.01) and a favorable shift toward lower disability across the 12-month mRS distribution (common OR for favorable shift 3.57, 95% CI 1.42–8.93; p<0.01). Cumulative 12-month mortality was similar between groups (both 18%). Among survivors, 6% of the Cerebrolysin group versus 19% of controls required institutional care (unadjusted OR 0.26; 95% CI 0.07–0.99; NNT 8). BI scores were higher in the Cerebrolysin group than in controls (median (Q1–Q3) 92 (82–100) vs 83 (73–93); p=0.01). In multivariable models, Cerebrolysin remained associated with 12-month independence alongside complete reperfusion (mTICI 3), lower post-EVT NIHSS, fewer device passes, and absence of symptomatic intracranial hemorrhage. Conclusions In EVT-treated patients selected for a small infarct core, robust collaterals, and high-quality reperfusion, adjunctive Cerebrolysin showed a potential benefit toward better 12-month functional outcomes. These exploratory findings require confirmation in multicenter randomized trials to establish efficacy and refine patient selection
Comparing human vs. machine-assisted analysis to develop a new approach for Big Qualitative Data Analysis
Get PDFBackground: The exponential growth of Big Qualitative (Big Qual) data in healthcare research presents methodological challenges for traditional analysis approaches. This study evaluates the effectiveness of machine-assisted analysis using artificial intelligence (AI) tools compared to human-only analysis for processing large-scale qualitative datasets, using the Royal College of Anaesthetists’ 7th National Audit Project (NAP7) baseline survey as a test case. Methodology/Principal Findings: We conducted a comparative methodological study analysing 5,196 free-text responses about peri-operative cardiac arrest experiences. Three researchers established a human-coded reference standard following SRQR guidelines. We then applied machine-assisted analysis using Pulsar for exploratory analysis and Caplena for sentiment and thematic analysis, evaluating performance against the human gold standard using STARD-AI reporting standards. Performance metrics included accuracy, precision, recall, F1-scores, and Cohen’s Kappa, with confidence intervals calculated using bootstrap resampling. Machine-assisted analysis substantially reduced analysis time, with particularly dramatic improvements in theme identification speed. The machine-assisted approach achieved good thematic and sentiment classification accuracy compared to the human reference standard, though human analysis identified an emergent ‘ambiguous’ sentiment category that current AI tools cannot accommodate, highlighting limitations in commercial platforms’ flexibility for inductive analysis. Conclusions/Significance: Machine-assisted analysis offers substantial efficiency gains with acceptable accuracy trade-offs for large-scale qualitative data analysis. However, human expertise remains essential for capturing nuanced meanings, identifying emergent categories, and providing domain-specific interpretation. This hybrid approach represents a viable methodology for Big Qual research, though current AI tools’ constraints in accommodating emergent classification schemes remain a limitation. Our findings establish benchmarks for future development of more flexible AI systems adapted to qualitative research paradigms
From subduction initiation to oceanic core complex formation: a tale of two ophiolites in the Kurdistan, NW Iran
Get PDFThe processes of subduction initiation and oceanic core complex (OCC) formation are essential to understanding plate tectonic evolution. The Kermanshah–Walash Ophiolitic Complex (KWOC) in Kurdistan, NW Iran, preserves a unique record of both mechanisms in spatially parallel but temporally distinct Late Cretaceous (Razab) and Eocene (Sarv-Abad) ophiolites. These ophiolites comprise peridotite, gabbroic bodies, a dike complex, and pillowed and massive basaltic lavas. Whole-rock geochemistry, U–Pb zircon geochronology, and field structures reveal distinct ages and tectonic settings for the two ophiolites. While radiolarian cherts constrain the Razab ophiolite to the Cretaceous, new U–Pb zircon (42–38 Ma) geochronology firmly places the formation of the Sarv-Abad oceanic core complex in the Eocene. The Cretaceous Razab unit contains high-Cr# spinel chromitite (0.77–0.82), forearc, and island arc tholeiitic (IAT) lavas, supporting formation during subduction initiation. In contrast, the Eocene Sarv-Abad unit features mylonitized gabbros, syn-extensional dikes, and contemporaneous D-MORB/E-MORB magmatism derived from heterogeneous mantle melting, diagnostic of an Oceanic Core Complex (OCC) in an extensional back-arc. Our results demonstrate that the KWOC was shaped by two distinct geodynamic mechanisms: subduction initiation in the Cretaceous and oceanic core complex formation in the Eocene. Despite ∼50 Myr separation, both units were emplaced during the Miocene closure, suggesting how contrasting mechanisms may have shaped the KWOC. The OCC model also offers a potential resolution to prior ambiguities in the geodynamic evolution of Peri-Arabic ophiolites
