179566 research outputs found
Sort by
In a Mirror, Dimly:Why AI Can’t Tell Our Stories and Why We Must
Get PDFToday’s generative AI tools are flooding the media ecosystem with mirrored reflections of humanity’s digitized past, reconstituted as the future. Companies are rapidly embracing these tools as ways to automate the already endangered professions of storytelling and knowledge creation. Why should we resist? After all, telling our own stories can often be painful and risky, frustrating and fruitless, or just tedious. What do we lose by surrendering the task of creating and conveying knowledge to machines that promise to remove the psychological, emotional and epistemological friction of storytelling? This article explores AI’s unwinding of the inextricable bonds among storytelling, human wisdom, knowledge and purpose, and why our future depends on their renewal
Assessing the efficacy, safety and utility of hybrid closed-loop glucose control compared to standard insulin therapy combined with continuous glucose monitoring in young people (≥16 years) and adults with cystic fibrosis related diabetes (CL4P-CF study): protocol for an open-label, multi-centre, randomised, two arm and single period parallel trial
Get PDFIntroduction Cystic fibrosis related diabetes (CFRD) is one of the most clinically impactful co-morbidities associated with cystic fibrosis. Current recommended management with insulin therapy is challenging due to variable daily insulin requirements and adds to the significant burden of self-management. This study aims to determine if hybrid closed-loop insulin delivery can improve glucose outcomes compared to standard insulin therapy with continuous glucose monitoring (CGM) in young people (≥16 years) and adults with CFRD.Methods and analysisThis open-label, multi-centre, randomised, two arm single-period parallel design study aims to randomise 114 young people (≥16 years) and adults with CFRD. Following a 2-3 week run-in period, during which time participants use a masked CGM, participants with time in target glucose range (3.9 to 10.0mmol/L) <80% will be randomised to 26 weeks with hybrid closed-loop insulin delivery or standard insulin therapy with CGM. The primary outcome is the between group difference in time in target glucose range (3.9 to 10.0mmol/L) based on CGM levels during the 26-week study phase. Analyses will be conducted on an intention-to-treat basis. Key secondary outcomes are time above target glucose range (>10.0mmol/L), mean glucose and HbA1c. Other secondary efficacy outcomes include glucose and insulin metrics, change in forced expiratory volume in 1 second (FEVI) and body mass index (BMI). Safety, utility, participant experiences and participant reported outcome measures (PROMs) will also be evaluated. The trial is funded by the National Institute of Health and Care Research (NIHR). Ethics and dissemination Ethics approval has been obtained from East of England - Cambridge South Research Ethics Committee. Results will be disseminated by peer-reviewed publications and conference presentations, and findings will be shared with people living with CF, healthcare providers and relevant stakeholders.Trial registration: NCT05562492 (ClinicalTrials.gov) – Last updated October 16, 2024<br/
Artificial intelligence driven clinical decision-support systems to assist healthcare professionals and people with diabetes at the point of care: a Delphi-based consensus roadmap
Get PDFBackground & aimsThe use of Artificial intelligence (AI) to improve the diagnosis, assessment and treatment of people with diabetes, has the potential to drive a paradigm shift in diabetes care, both minimizing treatment inertia and optimizing clinical outcomes. This is a significant opportunity, given the predicted increase in the burden of diabetes over the next 20 years. However, there are concerns that regulatory processes for development and implementation of AI-driven technologies are not adequate for systems that may adapt to new data and change from their original performance characteristics as evaluated.MethodologyThe European Diabetes Forum (EUDF) convened a working group to review and investigate the unmet needs for implementation of AI technology in diabetes care. Through a series of virtual and face to face meetings, including email conversations, this author group developed the framework and focus of the accompanying analysis over six separate manuscript drafts. At the conclusion of the process, a series of 14 recommendations was authored using the Delphi methodology, requiring a two-thirds majority to agree on statements that would appear in the final document.OutcomesThe working group examined the key objectives for good diabetes care in the context of current and predicted AI-driven clinical decision support systems (AI-CDSS), including the outcomes for people with diabetes, the goals for personalized medicine and the implications for guideline-driven diabetes services and healthcare professionals (HCPs). The process covered the needs of primary care healthcare professionals, who will shoulder the majority of diabetes care. The challenge of developing regulatory concepts and processes that are sufficiently robust to be AI-inclusive was considered as central to the outcomes.ConclusionsBased on the available evidence, the EUDF working group believe that AI-CDSS will deliver benefits for people with diabetes, although there are clear challenges for moving AI-CDSS into the practical clinical space. To encourage debate on how this can be achieved safely and effectively, a series of 14 recommendations was agreed, which are discussed in context in this publication.<br/
Effectiveness of the maternal RSVpre-F vaccine against severe disease in infants in Scotland, UK:national population-based case-control and cohort analyses
Get PDFBackground:Respiratory syncytial virus (RSV) is a leading cause of infant hospitalisation, particularly in those under six months. In 2024, Scotland introduced a maternal RSVpreF vaccination programme, offered from 28 weeks gestation. While clinical trials demonstrate high efficacy, this study assessed vaccine effectiveness (VE), providing real-world evidence to inform policy and programme delivery.Methods:We conducted a retrospective, nested case–control study of all infants aged ≤90 days in Scotland (n=27,565) between August 12, 2024 and March 31, 2025, using routinely collected clinical data to estimate the maternal RSV VE against, the outcome, RSV-related lower respiratory tract infection (LRTI) hospitalisations at ≤90 days. Cases were hospital admissions with an LRTI and a positive RSV polymerase chain reaction (PCR) test within 14 days before or two days after admission and were matched to 10 controls by birth week and gestational age. Individuals with a positive RSV test but no LRTI hospital admission or a LRTI hospital admission and no RSV positive test are not considered to have the outcome of interest. Individual-level mother-baby data were used to estimate VE: births were identified using Scottish Linked Pregnancy and Baby Dataset (SLiPBD), vaccination records were obtained from the National Clinical Data Store (NCDS) and RSV-related hospital admissions were determined using RSV positive tests, obtained from Electronic Communication of Surveillance Scotland (ECOSS), and hospital admissions, obtained from Scottish Morbidity Records (SMR01). VE was estimated using adjusted conditional logistic regression, calculated as 100 × (1−odds ratio).Findings:During the study period, 13,842/27,565 (50.2%) pregnant women received the vaccine. 354 infants aged ≤90 days were hospitalised with RSV, with 3,511 matched controls. Among cases, 61/354 (17.2%) were born to vaccinated mothers, compared to 1,713/3,511 (48.8%) among controls. The median week of gestation at vaccination was 30 weeks (IQR 28-33 weeks). VE against RSV-associated LRTI hospitalisation was 82.0% (95%CI: 74.7-87.1) in infants born to vaccinated mothers compared to those born to unvaccinated mothers. VE remained high among preterm infants (<37 weeks: 89.7%, 95%CI: 54.6-97.7). Sub-optimal immunisation (≤14 days between vaccination and birth) did not confer protection (28.6%, 95%CI: -20.5–57.7). Vaccination averted 219 (95%CI: 189-243) RSV-related LRTI hospitalisations in infants aged ≤90 days.Interpretation:This first national, population-based study, provides robust evidence that maternal RSV vaccination substantially reduces RSV-related hospitalisations in infants ≤90 days, including pre-term infants. Maternal RSV vaccination programmes should be scaled up globally, with high coverage, offering potential to avert many RSV-associated infant hospitalisations.<br/
Large-scale historical land use mapping in Vietnam and Laos using military topographic maps
Get PDFLand use cover change (LUCC) is a major driver of global environmental and socio-economic transformations, with implications for carbon emissions, biodiversity, and sustainable development. However, robust historical analyses have often been limited by a lack of high-quality, spatially detailed baseline data. This study addresses this gap by being the first to apply deep learning-based image segmentation techniques to extract a comprehensive set of land use cover (LUC) information from historical topographic maps at a fine spatial resolution and at a cross-country scale. Specifically, we utilized topographic maps (1:50000) created by the U.S. Army Map Service during the Vietnam War (1963–1973) to create detailed historical LUC maps of Laos and Vietnam. We compared multiple model architectures on the manually labeled training data, with the UNet++ achieving the best performance. The resulting maps, produced at 4 m and 30 m resolutions, include 10 LUC classes and achieved high overall accuracies of 98.8% for Laos and 98.6% for Vietnam on separate test sets. Analysis of the maps revealed forest cover losses of 18.2% in Laos and 25.0% in southern Vietnam (below 17° N) by 1990 and a 36.8% reduction of mangrove forests in Vietnam by 1996. Transitions from forest to shrubland dominated in northern parts of Laos and Vietnam while transitions to cropland were most prevalent in Savannakhet province of Laos and the Southeast region of Vietnam. The maps provide a novel baseline for assessing post-war LUCC dynamics in Southeast Asia allowing for spatially explicit analyses of conflict and policy impacts at the time. Furthermore, the study demonstrates a transferable methodology that makes archival map collections accessible for large-scale historical global change research
Patient and Public Involvement and Engagement in methodology research:Process, experiences, and recommendations from the SPIRIT- and CONSORT-Surrogate project
Get PDFBackgroundWhile there are increasing calls for Public and Patient Involvement and Engagement (PPIE) in methodology research, including the development of reporting guidelines, practices continue to emerge. This paper reports on the process, experiences, reflections, and recommendations of both the PPIE partners and other researchers participating in the development of (Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) and Consolidated Standards of Reporting Trials (CONSORT)-Surrogate reporting guidelines.MethodsDevelopment of the SPIRIT- and CONSORT-Surrogate guidelines involved four phases: (1) literature reviews; (2) an e-Delphi survey; (3) a consensus meeting, and (4) knowledge translation. PPIE was integrated in Phases 2, 3 and 4. An encompassing budgeted PPIE strategy detailing involvement in all project phases was prepared and implemented by researchers and PPIE partners. Implementation included a learning workshop (attended by 19 PPIE partners) to build PPIE partners’ capacity and confidence to participate in the e-Delphi survey (Phase 2) and the invitation of four PPIE partners to the consensus meeting (Phase 3). Experiences and reflections of PPIE in the project, based on feedback surveys from PPIE partners participating in the project and reflective notes from meetings, were used to formulate recommendations.ResultsIn total, 19 PPIE partners took part in the e-Delphi survey (Phase 2), four joined the consensus meeting (Phase 3), and consequently co-authored the guidelines and contributed to the development of an educational animation video (Phase 4). Partners felt that facilitators for involvement in Phase 2 included a learning workshop, financial compensation, support during e-Delphi survey participation (such as a glossary and help texts) and for Phase 3, the main facilitator was allowing partners to contribute first during the consensus meeting. The PPIE partners who joined the consensus meeting (Phase 3) presented the patient perspective; reminded researchers of why the project was important; helped with clarification of issues; corrected grammar; suggested strategies to disseminate and implement the extensions; and created humour. Reflecting on the involvement, both the PPIE partners and researchers felt it was valuable to the project.ConclusionsBased on the experiences, we make six recommendations for integrating PPIE in projects to develop reporting guidelines: involve early; involve with a plan and layered approach; involve meaningfully in a genuine way; involve with support and in safe spaces; involve with reflection and feedback; and involve with a budget to compensate for time and effort
Quantitative sensory testing reveals evidence of altered pain processing in Paget’s Disease of Bone
Get PDFBackgroundPain is the most common symptom of Paget’s disease of bone (PDB), but its underlying mechanisms are poorly understood. Notably, bone pain does not correlate well with metabolic activity or treatment response. This study aimed to assess whether sensory processing is altered in skin overlying Pagetic bone using quantitative sensory testing (QST).MethodsWe conducted a cross-sectional study of 156 people with PDB attending secondary care referral centres in the UK. We conducted quantitative sensory testing of the skin overlying affected sites and compared the data with unaffected sites as a control. The modalities used were hot and cold rollers, pinprick, vibration and von-Frey filaments to test both spinothalamic and lemniscal pathways.ResultsThere was a consistent trend for sensory perception to be increased over affected sites versus control sites in the study population. The differences were significant for vibration detection threshold (p = 0.009), pain threshold (p = 0.010) and both single and multiple pinprick testing methods (both p < 0.001). Subgroup analysis revealed similar trends when analysis was restricted to those with pain thought to be due to bone deformity or increased metabolic activity and those with and without musculoskeletal pain.ConclusionSensory processing is altered in skin overlying Pagetic bone, independent of current pain symptoms. We speculate that this may be due to abnormalities of bone shape, bone structure or metabolic abnormalities in the affected bone. The mechanisms are unclear but deserve further study
High throughput screening identifies activity of histone deacetylase inhibitors in patient-derived models of soft tissue sarcoma
Get PDFBACKGROUND: Undifferentiated pleomorphic sarcoma (UPS) is a rare and aggressive soft tissue sarcoma with limited treatment options and a poor prognosis. As a complex karyotype tumor, UPS lacks recurrent targetable mutations, and response rates to standard first-line doxorubicin therapy are low. Phenotypic drug screening offers an alternative approach to identify new therapeutic targets without requiring prior knowledge of molecular mechanisms.METHODS: A library of FDA-approved compounds and a custom histone deacetylase (HDAC) inhibitor library were screened using well-annotated patient-derived cell lines. Hit compounds were further characterized using apoptosis assays and in vivo xenograft studies. Biomarkers of activity were evaluated using gene expression and western blot analyses. Synergy with doxorubicin was evaluated in combination assays. RESULTS: HDAC inhibitors emerged as a promising therapeutic class, demonstrating low IC 50 values across cell lines (14.8-26.89 nM), with quisinostat taken forward for further evaluation. Gene expression changes in EPAS1, FOXO1, AMOT, and FOSL1 were observed as potential biomarkers of activity. Combination assays revealed synergy between quisinostat and doxorubicin (average ZIP score: 1.02-15.65; ZIP max: 3.98-33.71), increasing apoptotic cell death in vitro. In vivo, quisinostat alone and in combination with doxorubicin significantly reduced the tumor volume (vehicle 160.0 ± 63.2 mm 3, doxorubicin 78.0 ± 35.2 mm 3, quisinostat 84.3 ± 13.1 mm 3, and combination 49.2 ± 10.2 mm 3). Quisinostat also showed potent activity in leiomyosarcoma (LMS) cell lines (5.82-31.32 nM), which represent an additional complex karyotype soft tissue sarcoma. CONCLUSIONS: Quisinostat demonstrated strong preclinical activity and synergy with standard-of-care doxorubicin in models of UPS and LMS.</p
Alignment and anisotropy of stresses in disordered granular media
Get PDFCharacterizing the degeneracy of local stress states is a central challenge in obtaining the com-plete statistical mechanics of disordered media. Here, we introduce a minimal force-balance model for isolated granular clusters to probe the structure of the stress space through principal stress ori-entation and stress anisotropy. We further show that when complemented by physically motivated pairwise constraints, the model produces predictions for the stress alignment in packings of repul-sive hard spheres. We compare these predictions against simulation data for grains in hopper and simple shear flows, finding qualitative agreement. This demonstrates the promise of modeling bulk athermal disordered systems through the combinatorics of few primitive geometric motifs
The politics and profit of disinformation in public health
Get PDFDisinformation is a coordinated or deliberate effort to knowingly circulate misinformation (i.e., false information) to gain money, power, or reputation. While most public health research has focused on misinformation, disinformation can have particularly pernicious direct and indirect public health effects, including a growing impact on public health policy and the reputation of public health evidence and institutions. This review focuses on the role of disinformation for profit among multinational corporations, antiscience policy, and how and why disinformation is increasing. It also examines approaches to address disinformation in public health and social policy, such as a greater focus on the nature and power of framing, strategies for "prebunking" of predictable narratives, and denormalization and countermarketing.</p
Access Repository Dashboard
Do you manage Edinburgh Research Explorer? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!