Publikationer från Umeå universitet
Not a member yet
55455 research outputs found
Sort by
Public participation and grand narratives of constitutional transitions : the case of Fiji
No full textTheunis Roux’s article about two opposing grand narratives of constitutional transition,the liberal-progressive narrative and the culturalist-decolonial narrative, is a thought-provoking read. Though Roux makes an important contribution as he inspires us to bringthe two narratives into conversation with each other, his essay can also be criticised asa dichotomous interpretation that does not account for an empirical reality that is muchmore complex. Whether or not decolonized countries have constitutions that “reflect thevalues of the Westernised political elites that adopted them” is not the focus of this article,however. Rather, the focus here is to understand the process of making the constitution andparticularly so when the population at large are invited to participate; what has come tobe termed as “participatory constitution-making”. One might wonder how, then, does thisarticle relate to Roux’s ideas of the two different grand narratives, if it does not deal directlywith constitutional content? I argue that the notion of participatory constitution-makingforms part of a liberal-progressive narrative in the sense that the call for broad based participation is strongly advocated by primarily (western) international organizations. I haveelsewhere discussed that the extent to which contemporary constitution-making processeshave been participatory—in the sense of allowing participants to exert influence—widelyvaries between cases. Roux’s article, however, spurs additional thoughts on why certaincases have talked the talk of “participation” but not quite managed to walk the walk. In thispiece, I will focus on a case in the Pacific region—namely Fiji—to elaborate this matter.
A comparison of deep neural network compression for citizen-driven tick and mosquito surveillance
No full textCitizen science has emerged as an effective approach for infectious disease surveillance. With advancements in machine learning, entomologists can now be relieved from the labor-intensive task of species identification. However, deploying machine learning models on mobile devices presents challenges due to constraints on battery life and memory capacity. In this study, we explore the potential of various model compression techniques for deploying machine learning models on resource-limited devices, enabling low-energy consumption and on-device processing for disease surveillance in remote or low-resource settings. We compared two main-stream model compression techniques, pruning and quantization on various mobile devices. Our findings indicate that quantization methods outperform pruning methods in terms of efficiency. Furthermore, we propose to integrate structured and unstructured pruning to enhance model performance while addressing key constraints of mobile deployment
How to quickly select good in-context examples in large language models for data-to-text tasks?
No full textIn the realm of data-to-text generation tasks, the use of large language models (LLMs) has become common practice, yielding fluent and coherent outputs. Existing literature highlights that the quality of in-context examples significantly influences the empirical performance of these models, making the efficient selection of high-quality examples crucial. We hypothesize that the quality of these examples is primarily determined by two properties: their similarity to the input data and their diversity from one another. Based on this insight, we introduce a novel approach, Double Clustering-based In-Context Example Selection, specifically designed for data-to-text generation tasks. Our method involves two distinct clustering stages. The first stage aims to maximize the similarity between the in-context examples and the input data. The second stage ensures diversity among the selected in-context examples. Additionally, we have developed a batched generation method to enhance the token usage efficiency of LLMs. Experimental results demonstrate that, compared to traditional methods of selecting in-context learning samples, our approach significantly improves both time efficiency and token utilization while maintaining accuracy
Plasma proteins associated with cardiovascular disease in relation to lung function in SCAPIS
No full textBACKGROUND: Low lung function has been consistently associated with increased cardiovascular disease (CVD) risk, with emerging evidence suggesting a potential causal relationship. However, underlying biological mechanisms remain unclear. AIM: To investigate relationships between CVD-associated plasma proteins and lung function. METHODS: We analysed plasma protein profiles in two Swedish population-based cohorts: the Swedish CArdioPulmonary bioImage Study (SCAPIS) (n = 4,982, mean age 57.6 years) as the discovery cohort and the SCAPIS pilot study (n = 1,054, mean age 57.7 years) for replication. Multiple linear regression models were used to assess associations between 92 CVD-associated proteins and z-scores of FEV1, FVC, and FEV1/FVC, adjusting for known confounders. P-values were corrected using the Benjamini-Hochberg method (5% FDR). Significantly associated proteins were validated in the replication cohort. R ESULTS: A total of 69 proteins were associated with FEV1, 57 with FVC, and 9 with FEV1/FVC. Several inflammatory proteins and adipokines, including leptin, interleukin-6, fatty acid-binding protein (adipocyte), were consistently linked to lower lung function. Leptin had the strongest negative association (FEV1: β = -0.50, 95 % CI: [-0.69, -0.31], p < 0.001; FVC: β = -0.52, 95 % CI: [-0.68, -0.35], p < 0.001 per-SD increase). CONCLUSIONS: Multiple CVD-associated proteins, mainly reflecting inflammatory and metabolic processes, were associated with reduced FEV1 and FVC, supporting a link between systemic inflammation, adipokine metabolism and impaired lung function. Leptin had the strongest association, suggesting that its effects on lung function may extend beyond adiposity. Further research is needed to clarify the mechanisms driving these associations and to assess whether these proteins could serve as early biomarkers or intervention targets
Relationship between active aging and perceived aspects of home among persons aged 55+ listed with an interest in relocation
No full textThe objectives of this study were to explore active aging among people listed with an interest in relocation, and the relationships between active aging and aspects of perceived home. Using cross-sectional data from the 2022 RELOC-AGE project (N = 1,509, mean age = 70 years), linear regression analysis was conducted. Women and individuals with higher self-rated health and education reported higher levels of active aging. After adjusting for confounders, Housing Satisfaction (decrease in Housing satisfaction led to lower active aging scores (beta = -17.8, 95% CI [-28.6, -7.0] for neither satisfied nor dissatisfied), and Meaning of home relationship (beta = 4.6, 95% CI [3.4, 5.8]), was positively associated with active aging, "Housing-Related Control Beliefs" showed a significant negative relationship (beta = -14.6, 95% CI [-17.1, -12.1]). These findings are significant for promoting health and well-being among older adults. They add knowledge about home as a key factor for active aging and could be valuable for policymakers, housing authorities, and healthcare and social services staff involved in aging and housing issues
Legitimacy in Balance: Trategies and Role Ambiguity in the Professional Practice of School Counselors : School counselor experiences of psychosocial work in school
No full textDenna studie syftar till att belysa hur skolledningens styrning formar skolkuratorns arbete samt hur professionen navigerar i spänningsfältet mellan pedagogiska mål och psykosocialt arbete. Särskild uppmärksamhet i studien riktas mot hur organisatoriska villkor, institutionella logiker och legitimitetsskapande strategier påverkar kuratorns handlingsutrymme och möjligheter att bedriva ett förebyggande arbete. Undersökningen har genomförts med hjälp av kvalitativa semistrukturerade intervjuer med fem skolkuratorer verksamma inom kommunal grund- och gymnasieskola. Den insamlade empirin bearbetades genom en tematisk analys och tolkades utifrån ett nyinstitutionellt teoretiskt ramverk, med fokus på begrepp som isomorfism, särkoppling och institutionella logiker. Studien visar att skolkuratorns yrkesroll formas i en tydlig konflikt mellan skolans pedagogiska logik och elevhälsans omsorgslogik, där akuta insatser systematiskt tenderar att tränga undan det lagstadgade förebyggande arbetet. Resultatet indikerar att organisatorisk placering är avgörande; en central placering fungerar som ett skyddande filter för professionens autonomi, medan en placering direkt under rektor ökar risken för anpassning och rollglidning. Vidare framkommer att kuratorer, i brist på en självklar institutionell legitimitet, använder mimetiska strategier och tar på sig arbetsuppgifter utanför sitt kärnuppdrag för att bygga förtroende och relationer i skolmiljön
High LRIG1 expression predicts lymph node metastasis in patients with uterine cervical cancer
No full textFifteen percent of patients with preoperative stage IA2-IB1 uterine cervical cancer are diagnosed with lymph node metastasis (LNM) following surgery. They must be treated with both surgery and radiotherapy, a combination associated with severe side effects. Since current diagnostic methods have limitations, biomarkers are urgently needed to improve staging. Leucine-rich repeats and immunoglobulin-like domains protein 1 (LRIG1) is a regulator of growth factor signaling and a prognostic factor in cervical cancer. This study investigates whether LRIG1 expression could predict LNM in cervical cancer. Sixty-seven patients were included: 31 without LNM and 36 with LNM. Tumor blocks were retrieved, and clinical data were collected. Immunohistochemical analysis of LRIG1 expression was performed, and LRIG1 immunoreactivity was correlated with lymph node status and clinicopathological prognostic factors, such as human papillomavirus status and smoking status. High LRIG1 expression (> 25% positive cells) was significantly associated with an increased risk of LNM (odds ratio 9.49, 95% CI: 1.80-50.05, P = 0.008, adjusted for age, smoking status, and BMI), suggesting the potential of LRIG1 as a biomarker. Larger, multicenter studies are needed to validate our results
Heuristic search and constraint verification for value-centric electrification planning
No full textExpanding electrification infrastructure demands route planning tools that optimize feasibility while respecting legal, environmental, and stakeholder constraints. We present a hybrid system for value-centric electrification planning that combines heuristic search with logic-based verification. A modified A* algorithm explores alternatives using expert-informed heuristics, while an Answer Set Programming (ASP)-based rule set evaluates stakeholder constraint compliance. This separation of planning and evaluation supports transparency and explainability. Expert feedback highlights the system's value in early-stage decision support, and statistical analysis suggests that it consistently identifies technically feasible routes with reduced private land intrusion
Developing a calibrated physics-based digital twin for construction vehicles
No full textThis paper presents the development and calibration of a digital twin for a wheel loader that integrates a physical machine with a high-fidelity virtual model. The digital twin supports automated diagnostics, operational optimisation, and predictive simulations to enhance construction efficiency. Calibration using experimental data from the physical wheel loader improves the model’s accuracy, ensuring realistic replication of system mechanics. A physics-based multibody dynamics model was developed in AGX Dynamics as the core digital model. The physical loader was instrumented with pressure transducers on hydraulic cylinders, load pins to measure bucket forces in excavation, a quadrature encoder for linear displacement, and an inclinometer for bucket orientation. Data from actual operations were used to calibrate the digital model so that simulated excavation forces matched experimental measurements. Results show that while the uncalibrated model accurately predicts bucket loads before excavation, significant deviations occur during soil interactions. Calibration effectively mitigates these discrepancies, yielding a validated digital twin capable of accurate excavation-force prediction and reliable performance analysis.
Non-HDL and LDL cholesterol, but not calculated remnant cholesterol, are associated with subclinical atherosclerosis
No full textBACKGROUND: Elevated low-density lipoprotein (LDL) cholesterol levels represent a significant modifiable risk factor for atherosclerotic cardiovascular disease. However, a residual risk persists, possibly attributed to other atherogenic lipoproteins such as non-high-density lipoprotein (non-HDL) and remnant cholesterol. Nevertheless, few studies have explored the independent associations between these lipid biomarkers and early atherosclerotic disease. OBJECTIVE: To evaluate the relative contributions of LDL, non-HDL, and remnant cholesterol to subclinical atherosclerosis, assessed by carotid ultrasonography. METHOD: In this cross-sectional study, we included 1929 previously healthy individuals from the pragmatic VIPVIZA trial who had available lipid levels and carotid ultrasonography results to assess subclinical disease. Non-HDL, LDL, and remnant cholesterol were calculated from a standard lipid profile. Subclinical atherosclerosis was assessed by carotid intima-media thickness (cIMT) and the presence of carotid plaques. RESULTS: We found that all lipid variables (LDL, non-HDL, and remnant cholesterol) were associated with subclinical atherosclerosis in univariable models (P < .01 across all models for cIMT and P < .001, P < .001, P = .003 respectively for carotid plaques). In multivariable-adjusted models, increasing LDL and non-HDL cholesterol levels were still significantly associated with increased odds of having carotid plaques (P < .001 for both) and increased cIMT (P < .001 for both). However, no independent association between remnant cholesterol and subclinical atherosclerosis was observed in the model adjusted for LDL cholesterol levels (P = .073 for cIMT and = .818 for plaque). CONCLUSION: Increasing LDL and non-HDL cholesterol levels, but not remnant cholesterol, seem to contribute to carotid subclinical atherosclerosis