159370 research outputs found
Sort by
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
Energy efficiency support for software defined networks: a serverless computing approach
No full textAutomatic network management strategies have become paramount for meeting the needs of innovative real-time and data-intensive applications, such as those in the Internet of Things. However, the ever-growing and fluctuating demands for data and services in such applications require more than ever an efficient, scalable, and energy-aware network resource management. To address these challenges, this paper introduces a novel approach that leverages a modular architecture based on serverless functions within an energy-aware environment. By deploying SDN services as Functions as a Service (FaaS), the proposed approach enables dynamic, on-demand network function deployment, achieving significant cost and energy savings through fine-grained resource provisioning. Unlike previous monolithic SDN approaches, this work disaggregates SDN control plane into modular, serverless components, transforming tightly integrated functionalities into independent, on-demand services while ensuring performance, scalability, and energy efficiency. An analytical model is presented to approximate the service delivery time and power consumption, as well as an open source prototype implementation supported by an extensive experimental evaluation. Experimental results demonstrate significant improvement in energy efficiency compared to traditional approaches, highlighting the potential of this approach for sustainable network environments
Can the adoption of circular economy practices foster supply chain resilience and performance improvements?
Get PDFWhile a growing literature is showing interest in the circular economy (CE) paradigm, there is still a lack of consensus on whether the adoption of CE practices can help to cope with supply risks arising from an increasingly uncertain business environment in order to increase supply chain resilience (SCRES) and improve a firm's performance. Through a survey of Italian enterprises engaged with CE practices, this study aims to fill this literature gap, investigating whether the adoption of CE practices can initiate a path of increased SCRES, which can lead firms to improve their overall performance, thus proactively responding to environments characterised by high levels of supply risk. This study contributes to the debates about the paths connecting CE practices and firms' performance, especially in the context of vulnerabilities and disturbances, empirically demonstrating how firms might exploit the potential of CE by investing in SCRES. This study sheds light on the relationship between CE and SCRES, particularly underlying the most relevant paths of relationships between CE and those SCRES capabilities that can lead to performance improvements, particularly when the level of supply risk increases
The dimension of well approximable numbers
No full textIn this survey article, we explore a central theme in Diophantine approximation inspired by a celebrated result of Besicovitch on the Hausdorff dimension of well approximable real numbers. We outline some of the key developments stemming from Besicovitch's result, with a focus on the Mass Transference Principle, Ubiquity and Diophantine approximation on manifolds and fractals. We highlight the subtle yet profound connections between number theory and fractal geometry, and discuss several open problems at their intersection
The recovery of public sector accounting as a site of possibility: publicness and localized-led development
Get PDFAs public sector accounting stands at a momentous historical juncture, this editorial introduces a special issue of Financial Accountability and Management that reimagines public sector accounting in the wake of New Public Management's (NPM) limitations. Confronting crises of governance, legitimacy, and representation, we propose two generative analytical anchors—publicness and localized-led development—to challenge dominant managerial logics and reorient accounting toward more democratic, situated, and socially responsive practices. Publicness is reconceptualized as a dynamic, contested space shaped by accounting technologies, civic engagement, and political struggle. Localized-led development foregrounds the agency of place-based actors and vernacular knowledge in resisting technocratic reforms and enabling contextually grounded governance. The issue features five empirical studies spanning student unions in Australia, municipal services in Turkey, environmental activism in Malaysia, infrastructure failure in Latin America, and peacebuilding in Palestine. Together, they illuminate how accounting mediates between global reform discourses and local governance realities, functioning both as a disciplinary tool and a potential site of transformation. We argue that public sector accounting must move beyond critique and engage in reconstructive scholarship that values pluralism, participatory accountability, and care ethics. By foregrounding the political, cultural, and ethical dimensions of accounting, this issue offers a critical path forward for scholars and practitioners seeking to reclaim public sector accounting as a force for public value, justice, and sustainable development
Augmentation of immunothrombosis as a key mechanism underlying JAK inhibition associated hypercoagulability in rheumatoid arthritis
No full textObjectives
Venous thromboembolism (VTE), following Janus kinase inhibitors treatment (JAKi), is poorly understood in rheumatoid arthritis (RA). We investigated whether JAKi augmented immune cell-driven clotting or immunothrombosis in RA.
Methods
Peripheral blood leukocytes (PBLs) isolated from patients with RA and healthy controls were treated with various JAKi classes, before stimulation with Toll-like receptor (TLR)-4 (lipopolysaccharide [LPS]) or TLR3 (polyinosinic-polycytidylic acid—poly (I:C)) agonists. Conditioned supernatants were used in plasma turbidity assays to evaluate clot formation and lysis dynamics, while bulk RNA sequencing, enzyme-linked immunosorbent assay, and bead-based immunoassays were used to explore immunothrombosis mechanisms.
Results
Turbidity analyses showed that conditioned media from PBLs treated with LPS and tofacitinib significantly accelerated clot formation when compared to LPS alone, and this effect was tissue factor pathway dependent and accompanied by elevations in immunothrombotic cytokines, including tumour necrosis factor α, interleukin (IL)-1β, and IL-6. PBLs from patients with active RA exhibited significantly greater immunothrombotic potential compared to those with low disease activity, despite comparable baseline cytokine levels. RNA sequencing analysis revealed significant pathway enrichment in tofacitinib/LPS-treated PBLs, including activation of Nuclear Factor (NF)-κB pathways, increased tissue factor expression, and reduced levels of anticoagulant factors such as protein S. Pharmacological inhibition assays with 5 JAK therapies suggested that JAK1/tyrosine kinase 2-dependent effect underscored increased thrombosis but selective JAK3 inhibition did not reproduce the prothrombotic effects. Finally, patients with RA with JAK-associated pulmonary embolism showed interstitial changes compatible with immunothrombosis in 4/6 (67%).
Conclusions
Immunothrombosis offers a novel explanation for JAKi-associated VTE in RA
Injury and local injection and the risk of foot/ankle osteoarthritis: a case–control study in retired UK male professional footballers
Get PDFObjective
The objective of this study was to examine whether foot/ankle injury and injection contribute to the risk of foot/ankle OA in retired UK male professional footballers.
Methods
This was a case–control study among retired UK male footballers, in which cases reported General Practitioner–diagnosed foot/ankle OA or forefoot/ankle surgery after retirement, and controls reported neither. Injury was defined as significant foot/ankle injury with pain for most days over 3 months during their career. Injection was defined as injection of corticosteroids or other agents into foot/ankle joints during their career. Adjusted odds ratios (aORs) with 95% confidence interval (CIs) were calculated using logistic regression. Areas Under the Curve (AUCs) and 95% CIs were estimated to examine the contribution of injury and/or injection in the context of other available risk factors.
Results
Of 424 footballers studied, 63 had foot/ankle OA and 361 had neither. Cases had similar mean age (63.2 vs 63.0, P = 0.457) and BMI (27.7 vs 27.0, P = 0.240) to those of controls, but more foot/ankle injury (73.3% vs 42.5%, P < 0.001) and injections (75.0% vs 48.4%, P < 0.001), with aORs of 4.23 (95% CI 1.88–9.48) and 2.62 (95% CI 1.19–5.78), respectively. The AUC was 0.69 (95% CI 0.62–0.77) for injury, 0.74 (95% CI 0.66–0.81) for injury and injection, and 0.78 (95% CI 0.70–0.85) for all risk factors. Similar results were observed in footballers with ankle OA only.
Conclusion
Injury was a major risk factor for foot/ankle OA in retired UK male professional footballers. The role of injection needs cautious interpretation due to potential confounding by indication
A theory-based randomised controlled trial to increase delivery of behaviour change interventions by healthcare professionals
No full textBackground
Public health policies require healthcare professionals to incorporate health behaviour change interventions (HBCIs) into routine consultations. This study tested whether an "if-then" planning intervention could enhance HBCI delivery.
Methods
A randomised controlled trial involving 1008 UK NHS healthcare professionals compared an intervention group, who formed "if-then" plans, with an active control group. Data were collected at one, two, twelve, and thirteen months. Primary and secondary outcomes included the proportion of patients receiving HBCIs, time spent delivering HBCIs, and healthcare professionals’ perceived capabilities, opportunities, and motivations.
Results
The intervention group showed more sustained improvements in HBCI delivery over time compared to the control group, although the between-group difference at the final follow-up (T4) was not statistically significant. The intervention group significantly increased HBCI delivery between T1 and T2 (mean difference = 3.74; p = .009), and between T2 and T3 (mean difference = 4.45; p < .001), with delivery remaining higher at T4. The control group showed a significant increase only between T1 and T2 (mean difference = 8.79; p < .001). Statistically significant improvements were observed in psychological capability, reflective motivation, and automatic motivation to deliver HBCIs, particularly within the intervention group.
Discussion
The if-then planning intervention led to sustained improvements in HBCI delivery, with the intervention group showing significant increases between T1 and T2, and between T2 and T3, and maintaining higher delivery at T4. Although the final time point showed no significant between-group difference, findings support "if-then" planning as a practical strategy to integrate HBCIs into routine care
Predicting animal movement with deepSSF : A deep learning step selection framework
Get PDFPredictions of animal movement are vital for understanding and managing wild populations. However, the fine-scale, complex decision-making of animals can pose challenges for the accurate prediction of trajectories. Integrated step selection functions (iSSFs), a common tool for inferring relationships between animal movement and the environment, are also increasingly used to simulate animal trajectories for prediction. Although admitting a lot of flexibility, the iSSF framework is limited to its reliance on pre-defined functional forms for fitting to data, and iSSFs that involve complex functional forms to model detailed processes can be prohibitively difficult to fit and interpret.
Here, we present deepSSF, an approach to fit and predict animal movement data using deep learning. The deepSSF approach replaces the log-linear model of an iSSF with a neural network architecture that receives multiple environmental layers and scalar values as inputs and outputs a single layer representing the next-step probability. We demonstrate an example deepSSF model, built in PyTorch, consisting of distinct but interacting habitat selection and movement subnetworks. This allows for explicit representation of both selection and movement processes, thus giving interpretable intermediate outputs. We apply our model to GPS data of introduced water buffalo (Bubalus bubalis) in the tropical savannas of Northern Australia.
Our deepSSF model was able to learn features that are present in the habitat covariate layers, such as linear features (rivers, forest edges) and the composition of certain habitat areas, without having to specify them pre-emptively within the model framework. It was able to capture complex interactions between the habitat covariates as well as temporal dynamics across time of day and year. Finally, our deepSSF model generally had better in- and out-of-sample predictive accuracy than the analogous iSSF model.
We expect that the deepSSF approach will generate accurate and informative predictions about animal movement, which can be used for deepening our understanding of animal–environment systems and for the practical management of species. We discuss how the wide range of existing deep learning tools could enable the deepSSF approach to be extended to represent memory and social dynamic processes, with the potential for integrating non-spatial data sources such as accelerometers and physiological sensors
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇