Gustavus Adolphus College Collections
Not a member yet
92910 research outputs found
Sort by
Can I trust GenAI to plan my next trip? A multi-method approach to optimizing media mix
No full textAlthough tourists can now book trips directly using generative artificial intelligence (GenAI), it remains unclear whether the real-time travel information it provides is comprehensive and sufficiently trustworthy enough to make booking decisions. The present research addresses this gap by integrating media richness, trust transfer, and the value-based adoption model (VAM) to investigate the impact of varying levels of travel information richness (text-only, text-image, and text-image-audio) on the booking behaviors of tourists using GenAI such as ChatGPT. With data from 578 participants, we tested the proposed structural and configurational models using a multi-analytical approach. Our findings revealed that the three media richness levels yield both analogous and distinctive effects on tourist perceptions regarding benefits, costs, trust formation, and intentions in ChatGPT online travel booking. Specifically, the text-image group demonstrated the strongest links from media richness to trust in ChatGPT, perceived benefit to value, and ultimately value to increased booking intention. Our findings from configurational modeling confirm a significant opportunity to harness the power of AI-empowered platforms for online travel booking
Explainable colon cancer stage prediction with multimodal biodata through the attention-based transformer and squeeze-excitation framework
No full textIntroduction: The heterogeneity in tumours poses significant challenges to the accurate prediction of cancer stages, necessitating the expertise of highly trained medical professionals for diagnosis. Over the past decade, the integration of deep learning into medical diagnostics, particularly for predicting cancer stages, has been hindered by the black-box nature of these algorithms, which complicates the interpretation of their decision-making processes.
Method: This study seeks to mitigate these issues by leveraging the complementary attributes found within functional genomics datasets (including mRNA, miRNA, and DNA methylation) and stained histopathology images. We introduced the Extended Squeeze- and-Excitation Multiheaded Attention (ESEMA) model, designed to harness these modalities. This model efficiently integrates and enhances the multimodal features, capturing biologically pertinent patterns that improve both the accuracy and interpretability of cancer stage predictions.
Result: Our findings demonstrate that the explainable classifier utilised the salient features of the multimodal data to achieve an area under the curve (AUC) of 0.9985, significantly surpassing the baseline AUCs of 0.8676 for images and 0.995 for genomic data.
Conclusion: Furthermore, the extracted genomics features were the most relevant for cancer stage prediction, suggesting that these identified genes are promising targets for further clinical investigation
Clinical effectiveness of early surgical fixation versus cast immobilization for adults with a scaphoid waist fracture:five-year follow-up of the Scaphoid Waist Internal Fixation for Fractures Trial
No full textAIMS: In the Scaphoid Waist Internal Fixation for Fractures Trial (SWIFFT), surgical fixation was compared with cast immobilization using the Patient-Rated Wrist Evaluation (PRWE) score at one year as the primary endpoint. The aim of this longer-term study was to assess the clinical effectiveness of the two treatment pathways at a five-year follow-up. METHODS: Patients who had not withdrawn from the trial were invited for a clinical assessment at five years after randomization and asked to complete a patient-reported questionnaire in the clinic, or by post or telephone. The PRWE was collected along with the grip strength, range of motion (ROM), imaging data, EuroQol five-dimension three-level questionnaire (EQ-5D-3L) score, and the use of health resources. The repeated measures analysis model used at one year was repeated to include the five-year timepoint. RESULTS: A total of 344 patients provided a valid PRWE score at five years (78.4% of the original cohort). There was no significant difference between the groups at five years (mean difference 0.6 (95% CI -2.4 to 3.6); p = 0.709). These results were robust to sensitivity analyses, restricting data collection windows. Neither the pain and function subscales of the PRWE score nor the overall estimate of the treatment effect showed a significant difference between the groups. Neither non-compliance with allocation of treatment, the displacement of the fracture, nor the patient's preferred treatment at baseline had an effect on the PRWE scores between groups. Only seven patients for whom imaging was collected at five years had a nonunion, three of 146 (2.1%) in the fixation group and four of 121 (3.3%) in the immobilization group. Grip strength and ROM were similar between the groups. CONCLUSION: The clinical effectiveness of the two forms of treatment remained similar at five years. The recommendation that adult patients with a fracture of the waist of the scaphoid which is displaced by ≤ 2 mm should be treated inially with immobilization in a cast, followed by early fixation of a nonunion, is further corroborated by these findings
Is generative AI reshaping academic practices worldwide? A survey of adoption, benefits, and concerns
No full textAlthough generative AI is transforming academic research and education, little is known about the role, gender, international, and disciplinary variations in uptake and use. This 20-country survey of publishing academics shows the widespread awareness and adoption of generative AI tools in academia, but with substantial international and disciplinary differences, and some role and gender differences. In particular, females were 10 % less likely to use Gen AI frequently (daily or weekly) for research, which may exacerbate gender inequalities. Perhaps surprisingly, the highest adoption rates occurred in some non-Western nations, possibly because of a greater need for translation services. The highest awareness is in the social sciences, perhaps because of the greater need for text analysis. Across all groups, these tools were mainly used for academic writing rather than data analysis and support for critical thinking. Despite this, personalized instruction and problem-solving are among generative AI's most generally claimed benefits. However, participants in all groups were skeptical about the creativity, accuracy, and consistency of AI-generated content in academic contexts. The most significant concerns about using generative AI in academia were inaccuracy, plagiarism, discouraging critical thinking, a lack of transparency and explainability, intellectual property rights violations, and data privacy risks. For policymakers, the findings point to fields and countries that may need action to prevent falling behind, as well as the ongoing need to investigate and monitor the impacts of generative AI on research practices
The role of human capital and Industry 4.0 in socio-technical dynamic capabilities for freight transport resilience
No full textThis study examines the role of socio-technical dynamic resources and capabilities in enhancing the resilience of freight transport firms. Focusing on unique characteristics of freight transport sector, this study considers human capital and Industry 4.0 as socio-technical resources whereas network integration and operations planning as socio-technical capabilities. A survey is conducted on freight transport managers in the UK and partial least squares structural equation modelling (PLS-SEM) is employed to analyse the data. Our findings indicate that human capital and Industry 4.0 positively affect network integration and operations planning, which in turn enhance perceived resilience. Among two dynamic capabilities, network integration has a stronger impact on resilience than operations planning. Mediation analysis shows technical resources and capabilities partially enhance the impact of social resources and capabilities in SEM model. Our results suggest that socio resources and capabilities are at least as important as technical ones for building resilience in the freight sector. Practical and policy implications in our study suggest that freight transport firms and policy makers should invest and support skill-building initiatives, digital tools, supply chain analytics and collaboration platforms to increase freight transport resilience
Spatially resolved single-cell analysis of transcriptomic changes linked with neuropathic pain in human neuromas
No full textInjuries to the trigeminal nerve, responsible for sensory innervation to the face, may occur during routine dental procedures, resulting in the formation of a neuroma accompanied by loss of sensation and/or symptoms of pain. To gain insight into the molecular mechanisms underpinning the sensory changes, single nuclei RNA sequencing and spatial transcriptomics were used to profile the transcriptional landscape at single cell resolution of human trigeminal nerves and neuromas. Cellular and transcriptional changes were identified that correlated with the presence of pain, including an expansion of endothelial cells with a proinflammatory phenotype and overexpression of HLA-A, CXCL2, and CXCL8. Interactome analysis highlighted signaling changes linked with the presence of pain. HLA-A protein expression was confirmed in neuromas and positively correlated with symptoms of pain. Our data provide a detailed spatial overview of the cell types that populate human peripheral trigeminal nerves, in health and injury, and their transcriptional profile. The comparison of painful and nonpainful samples highlights several changes in cellular composition, transcriptome and inferred molecular signaling linked specifically with the presence of pain, and also a novel role for HLA-A in neuropathic pain. Our findings represent a valuable resource for pain research, highlighting the role of inflammation, endothelial cell dysfunction, and chemokine signaling in neuropathic pain
Environmental challenges in electrification: Traffic-induced non-exhaust PM2.5 emissions in Cangzhou, China
No full textNon-exhaust PM2.5 emissions are often overlooked despite their increasingly significant contribution. Including non-exhaust emissions to establish a complete road emissions inventory is critical for formulating effective control strategies and reducing health risks especially during the fleet electrification process. This study developed a PM2.5 emission estimation framework incorporating exhaust and non-exhaust sources, validated by roadside PM2.5 concentrations from a mobile cruise sensing system in Cangzhou, China. The results found that traffic-related PM2.5 concentrations contributing to roadside environments exhibit hourly variations, with an average proportion of 44.05 %. Analysis of peak-hour PM2.5 emissions revealed the non-exhaust sources dominate (91.38 %). Under scenarios with 35 %, 50 %, 75 %, and 100 % electrification rates, PM2.5 emissions will decrease by 6.96 %, 11.01 %, 17.77 %, and 24.53 %, respectively, compared to the baseline. This study highlights that fleet electrification will not completely eliminate PM2.5 emissions, and helps identify priorities and develop targeted emission control policies for medium-sized populous cities in developing countries
Sex Disparities in Ischemic Heart Disease Mortality Across United States: The Southern Burden of Excess Body Weight and Diet
No full textBackground
While prior studies have assessed sex disparities in ischemic heart disease (IHD) outcomes in the United States (US), little is known about how these differences vary across states.
Methods
We analyzed Global Burden of Disease (GBD) data from 2011 to 2021 for all 50 US states. Sex‐specific age‐standardized IHD mortality rates (ASMRs) and age‐standardized IHD prevalence rates (ASPRs) were used to calculate the mortality‐to‐prevalence ratios (MPRs) per 100,000 inhabitants per year, enabling comparison of death rates relative to the population at risk. Hawaii, the state with the lowest ASPR and ASMR in both sexes, served as the reference for Z‐score analyses, with 99% (Z > 2.58) or 95% (Z > 1.96) confidence thresholds.
Results
In 2011, 30 states had significantly higher MPRs in females than males (Z > 2.58), indicating elevated mortality in females with IHD. By 2021, only Arkansas (6.7% [UI: 4.6–9.3]) and Mississippi (7.1% [UI: 4.8–9.7]) exceeded this threshold, while 11 additional states, mainly in the South, retained moderate disparities (Z > 1.96). Higher MPRs were inversely associated with state GDP per capita. High body max index, processed meat intake, low fiber, and low vegetable intake emerged as the most significant contributors (Z > 1.96, 95%) to excess IHD mortality in females but not males.
Conclusions
Despite an overall decline, sex disparities in IHD mortality persist with Arkansas and Mississippi continuing to bear the highest burden. These disparities are largely driven by overweight/obesity and dietary risk factors, warranting targeted, state‐level interventions
The NICE experience of designing and utilising severity weights
No full textBackground
In January 2022, NICE changed from “End of Life” (EoL) to “severity” weights, whereby additional value is applied within cost effectiveness analysis to health benefits arising from health technologies deemed to qualify This study examines the relationship between these concepts, how they relate to patient age, and whether the new system is cost-neutral as intended.
Methods
Data was extracted from 555 NICE Technology Appraisal decisions from 2009 to 2024. Absolute (AS) and proportional shortfall (PS) severity indicators were estimated for pre 2022 decisions. Post 2022 decisions were judged against EoL criteria.
We describe the relationship between severity weights, including the constituent AS and PS elements, age and EoL. Comparisons are made between the distribution of AS, PS and overall severity categories using descriptive statistics and significance tests.
Results
AS and PS have different relationships with patient age. In NICE appraisals, AS reduces with age but the relationship is flat between 40 and 60 years. All decisions in the highest AS category (AS > 18 QALYs) have a starting age below 20 years. PS peaks around 60 years. EoL applies almost exclusively where age exceeds 40 years. 91 % of appraisal decisions meeting EoL would receive a severity weight above 1.
There is no difference in the mean severity weight between pre and post 2022 appraisal decisions (1.116 vs 1.125). Mean AS is lower in post 2022 appraisals.
Conclusions
Severity weights do not correlate precisely with EoL. They have been applied as often as expected. The change from EoL to severity weights has been approximately cost-neutral
Palindrome-mediated 16p13.3 triplications cause a recognizable neurodegenerative disorder with ataxia
No full textComplex neurodegenerative conditions have occasionally been associated with copy-number gains. Using microarray and genome sequencing on DNA samples from eleven individuals from nine unrelated families, we show that copy-number gains at 16p13.3 cause a severe, recognizable disorder characterized by early-onset progressive ataxia and cognitive decline (9–32 years). Most affected individuals also displayed peripheral neuropathy and scoliosis. Optic atrophy, nystagmus, and dystonia were more variable features. The neuroradiological phenotype comprises a distinctive combination of atrophy of the cerebellum and caudate nuclei. Co-segregation data showed that the structural variant (SV) had occurred de novo in 5 individuals and, for one other individual, had been inherited from a mosaic, unaffected parent. Triplicated segments of 16p13.3 were identified within the duplications. Although these varied in size (30–811 kb), the minimal region of overlap included a single gene (ATP6V0C) that is highly expressed in the cerebellum. RNA sequencing (RNA-seq) using whole-blood and fibroblast/lymphoblast cultures indicated increased expression of several genes within the SV, with ATP6V0C showing the most significant increase (up to 4-fold). In most cases, the central segment of the SV was proven to be inverted and lay immediately distal to a 144 kb palindrome. Across 500,000 individuals from the UK Biobank, we identified 19 duplications but no triplications at this locus. Further analysis of the consequences of ATP6V0C overexpression on the stoichiometry within the vacuolar H+-ATPase heteromer and on neurological function will provide valuable pathomechanistic insights. Together, our findings define palindrome-mediated triplication on 16p13.3 as the cause of a clinically distinct childhood-onset neurodegenerative disorder