Linnaeus University
Not a member yet
43313 research outputs found
Sort by
The impacts of climate and the extreme drought in 2018 on population growth in Swedish moth species
No full textInsects are pivotal to ecosystem diversity and functionality, yet they face increasing threats from anthropogenic climate change impacts. A growing body of studies reports the effects of changing temperature and precipitation patterns, but relatively few studies focus on the consequences for insect populations because of extreme weather events. Here, we examine population growth responses to temperature, precipitation and the extreme summer drought in Sweden in 2018. For this purpose, we used longitudinal data for 54 moth species collected between 2005 and 2023 using light traps at three sites in southeastern Sweden. We found a positive relationship between temperature and population growth rates across all study sites, while precipitation showed a positive relationship at two sites and no effect at the third. The results indicated a negative time-lag effect on population growth, at two of the sites, of precipitation the previous year, while there were no significant effects of temperature the previous year. Despite the extreme drought in 2018, moth populations remained resilient, with no dramatic decline in population growth between 2018 and 2019. Our results contrast earlier studies reporting severe declines in population growth in response to extreme drought events. The discrepancy may reflect a combination of region-specific effects of extreme weather events and that selected species in this study predominantly consist of range-expanding and migratory species, better able to withstand adverse conditions due to a higher climatic tolerance and being habitat and food plant generalists. Our findings underscore the importance of an increased knowledge of site-specific responses and effects of extreme weather events such as droughts when outlining conservation efforts
Health care pathways described by care-seekers following a call to the emergency medical communication center - A Swedish perspective
No full textIntroduction A call to an emergency number is often the first course of action among care-seekers seeking emergency care. The emergency medical communication centre assesses the call and dispatches resources to meet the care-seekers' care needs. Knowledge concerning specific healthcare pathways is sparse and does not include the care-seeker perspective. Therefore, the aim of this study was to explore care-seekers' experiences of healthcare pathways following a call to the national emergency number.Methods The study was an explorative cross-sectional study using data from telephone interviews with callers to the Swedish national emergency number. Study informants (n = 141) provided descriptions of incidents (n = 156) that led up to the call, as well as their experiences following a healthcare pathway. Data analysis included mapping healthcare pathways, descriptive statistics, and gathering descriptions of care-seeker experiences.Results and conclusions This study shows the multiple healthcare pathways for care-seekers following a call to the national emergency number. The accessibility and availability of the healthcare system are found to be difficult to grasp among care-seekers. Decisions on which pathway to follow are primarily made by healthcare professionals rather than the care-seekers, which supports the notion that the system is not being designed from the care-seekers' perspective. Expressions of increased vulnerability, being unsafe and feeling lost were common among the care-seekers, regardless of which pathway was followed. The findings of this study could be used as an incentive for healthcare providers to develop a healthcare system based on the found novel knowledge of the care-seekers' perspective
Potentialen hos AI för förutsägelse av läkemedelsbehandling vid depression : En scoping review
No full textDepression är en av de mest förekommande psykiska sjukdomarna globalt och kan ha en betydande påverkan på individens livskvalitet. Trots att flera olika läkemedel finns tillgängliga, svarar en stor andel av patienterna inte på det första behandlingsvalet, vilket ofta leder till en process av att prova olika läkemedel för att hitta bästa alternativ. Denna process förlänger lidandet för patienten men innebär också en belastning för vården och en ekonomisk belastning för samhälle. Under de senaste åren har intresset ökat för att använda artificiell intelligens för att förutsäga behandlingssvar vid depression. Detta skulle kunna skapa möjlighet att ge en mer individanpassad läkemedelsbehandling till patienter med depression. Syftet med denna scoping review är att sammanställa aktuell litteratur om hur väl artificiell intelligens kan förutsäga behandlingssvar på läkemedel av depression. Fokus ligger på att analysera vilka typer av data och algoritmer som använts samt vilken prediktionsnoggrannhet modellerna uppnått. Genom att kartlägga det aktuella kunskapsläget kan detta arbete ge en samlad bild av artificiell intelligens nuvarande potential inom individanpassad läkemedelsbehandling vid depression. Flera studier har visat att maskininlärning har potential att förbättra förutsägelser av behandlingssvar vid depression. Metoder där EEG-data använts har visat en särskilt hög prediktionsnoggrannhet där vissa studier kunnat visa över 80 % träffsäkerhet. Kombinationer av flera datatyper, såsom kliniska, molekylära och neuroimaging data kan i vissa fall förbättra prediktionsförmågan, medan andra studier visat att flera olika typer av data inte alltid leder till ett bättre resultat. Framgången ser ut att bero på vald datatyp, algoritm och hur data bearbetas. Fortsatt forskning krävs för att öka modellernas förmåga att särskilja läkemedelseffekter från placebo och andra icke-farmakologiska faktorer för att möjliggöra att tekniken kan användas kliniskt.Abstract Depression is one of the most common mental disorders globally and can significantly impact an individual’s quality of life. Despite the availability of various medications, a large proportion of patients do not respond to the first treatment choice, often leading to a trial-and-error process to find the best option. This process prolongs the patient’s suffering, places a burden on healthcare and imposes an economic strain on society. In recent years, there has been growing interest in using AI to predict treatment response in depression. This could enable more personalized medication treatment for patients with depression. The aim of this scoping review is to summarize the current literature on how well artificial intelligence can predict medication treatment response in depression. The focus is on analyzing the types of data and algorithms used, as well as the predictive accuracy achieved by the models. By mapping the current state of knowledge, this work aims to provide an overview of the present potential of artificial intelligence in personalized pharmacological treatment for depression. Several studies have shown that machine learning has the potential to improve predictions of treatment response in depression. Methods that use EEG data have demonstrated particularly high predictive accuracy, with some studies reporting over 80 % accuracy. Combining multiple data types, such as clinical, molecular and neuroimaging data can in some cases enhance predictive performance, while other studies have shown that using various types of data does not always lead to better results. The success appears to depend on the chosen data type, algorithm and how the data is processed. Further research is required to improve the model’s ability to distinguish drug effects from placebo and other non-pharmacological factors to enable technology to be used clinically
Fine-tuning Clinical Language Models to Identify Adverse Drug Events in Clinical Text : Machine Learning Approach
No full textBackground: Medications are essential for health care but can cause adverse drug events (ADEs), which are harmful and sometimes fatal. Detecting ADEs is a challenging task because they are often not documented in the structured data of electronic health records (EHRs) . There is a need for automatically extracting ADE-related information from clinical notes, as manual review is labor-intensive and time-consuming. Objectives: This study aims to fine-tune the pre-trained clinical language model, SweDeClin-BERT, for medical named entity recognition (NER) and relation extraction (RE) tasks, and to implement an integrated NER-RE approach to more effectively identify ADEs in clinical notes from clinical units in Sweden. The performance of this approach is compared to our previous machine learning method, which utilized conditional random fields (CRFs) and Random Forest (RF). Data Sources: A subset of clinical notes from the Stockholm EPR (Electronic Patient Record) Corpus, dated 2009-2010, containing suspected ADEs based on ICD-10 codes in the A.1/A.2 categories was randomly sampled. These notes were annotated by a physician with ADE-related entities and relations following the ADE annotation guidelines. Methods: We fine-tuned the SweDeClin-BERT model for the NER and RE tasks and implemented an integrated NER-RE pipeline to extract entities and relationships from clinical notes. The models were evaluated using 395 clinical notes from clinical units in Sweden. The NER-RE pipeline was then applied to classify the clinical notes as containing or not containing ADEs. Additionally, we conducted an error analysis to better understand the model’s behavior and to identify potential areas for improvement. Results: In total 62% of notes contained an explicit description of an ADE, indicating that an ADE-related ICD-10 code alone does not ensure detailed event documentation. The fine-tuned SweDeClin-BERT model achieved an F1-score of 0.845 for NER and 0.81 for RE task, outperforming the baseline models (CRFs for NER and Random Forests for RE). In particular, the RE task showed a 53% improvement in macro-average F1-score compared to the baseline. The integrated NER-RE pipeline achieved an overall F1-score of 0.81. Conclusions: Utilizing a domain-specific language model like SweDeClin-BERT for detecting ADEs in clinical notes demonstrates improved classification performance (0.77 in strict and 0.81 in relaxed mode) compared to conventional machine learning models like CRFs and RF. The proposed fine-tuned ADE model requires further refinement and evaluation on annotated clinical notes from another hospital to evaluate the model’s generalizability. In addition, the annotation guidelines should be revised, as there is an overlap of words between the Finding and Disorder entity categories, which were not consistently distinguished by the annotators. Furthermore, future work should address the handling of compound words and split entities to better capture context in the Swedish language. NLMED seed project funding by DIS
How Feedback is Provided in Biology Education : Teachers’ Use of Feedback in Teaching and Students’ Perceptions of Receiving Feedback: A Didactic Study
No full textÅterkoppling med fokus på formativ bedömning och synliggjort lärande lyfts i styrdokument och forskning som en viktig faktor för elevers lärande och utveckling. Dock visar tidigare studier att återkopplingens tillämpning i klassrummet inte är framåtsyftande. Syftet med den här studien är att undersöka hur återkoppling används i biologiundervisningen på gymnasiet och studerades utifrån lärarnas och elevernas perspektiv. Studien bygger på intervjuer med tre lärare samt en enkätundersökning där 36 elever svarade. Resultatet påvisar att muntlig återkoppling är den mest förekommande formen i den dagliga undervisningen. Däremot vid skriftliga inlämningar är skriftlig återkoppling vanligast, vid prov ärbåde skriftlig återkoppling vanlig i någon form samt muntlig genomgång individuellt med eleverna. Återkopplingens form och innehåll påverkas i hög grad av praktiska förutsättningar såsom tidsbrist och hur många kurser som läraren undervisar i. Det framkommer även att återkoppling sällan explicit kopplas till lärandemål, och att elevernas engagemang i kamratbedömning är begränsad. Slutsatsen är att det finns potential att vidareutveckla återkopplingsarbetet. Det kan särskilt göras genom att tydligare koppla återkopplingen till lärandemål och stödja elevernas metakognitiva och självreglerande förmågor.
The challenges of democracy : A case study of democratic resilience versus democratic backsliding in Poland and Hungary
No full textThe aim of this study is to examine the democratic development of Poland and Hungary Between 2010 and 2024. This with a focus on the theoretical frameworks of democratic backsliding and democratic resilience. Poland and Hungary share similar institutional and historical backgrounds as they both are post-communist states and EU-members. Their democratization wasaround the same time around 1990; however, their recent political trajectories have been quite different. The study applies a comparative case study method with a most-similar systems design and it is mainly based on the Varieties of Democracy (V-Dem) dataset. The findings reveal a clear case of sustained democratic backsliding in Hungary, marked by the systematic weakening of checks and balances, judicial independence, and media freedom, leading to its downgrade to Partly Free by Freedom House in 2019. Poland, while also experiencing democratic erosion—particularly after 2015 in the judiciary and rule of law—has shown signs of recovery since 2023 and has so far retained its Freestatus. This relative resilience is linked to a more active civil society, stronger local governance, and interventions by the European Union, which have curbed further authoritarian tendencies. The analysis shows that democratic backsliding provides the most consistent framework for explaining the case in Hungary while democratic resilience is useful in understanding the more mixed case of the trajectory in Poland but cannot independently explain its outcomes. Instead, resilience should be seen as a complementary concept, highlighting how certain institutions and actors can slow down or partially reverse erosion. The comparison further illustrates that while Hungary fell into the trap of executive power concentration, Poland managed to avoid a similar trajectory, albeit with vulnerabilities that remain
The points from Primrose Ridge : A possible Still Bay workshop on the Gauteng Grassland of South Africa
No full textWith this contribution we re-introduce a Middle Stone Age point assemblage from the Gauteng Grassland. Whilst the open-air site can no longer be excavated or dated, an updated interpretation of the Primrose Ridge points has the potential to shed new light on the Middle Stone Age occupation of the Witwatersrand. We apply a fine-grained approach, developed for compiling intra- and inter-assemblage comparative data, to assess how the Primrose Ridge points compare to some other Middle Stone Age point assemblages. We demonstrate that most of the points from Primrose Ridge conform to what can be expected from Still Bay point assemblages in terms of invasive bifacial retouch, lenticular cross sections and semi-circular or pointed butts as originally defined. In situ anvil boulders surrounded by fresh quartz chips recorded in the 1940s, and knapping mistakes provide reasons to think of Primrose Ridge as a workshop or production site for Still Bay-like quartz points. If our interpretation is correct, Primrose Ridge could have been used during the Still Bay phase – in terms of relative chronology – representing the first tentative record of this technocomplex on the Grassland Biome of interior South Africa.
Abstract Body : Willy Granqvist’s Kropp as Queer Abstraction, and Its Reception
No full textThe aim of the study (”Abstract Body – Willy Granqvist’s Kropp as Queer Abstraction, and Its Reception”) is to achieve a deeper appreciation of how sexuality is depicted in the Swedish poet Willy Granqvist’s (1948–1985) sixth poetry collection Kropp (Body) (1980). In the book, sexuality is presented as a prominent theme, but it is portrayed as elusive and diffuse, often depicted through multivalent poetic images. In this analysis sexuality in Kropp is examined, partly through a reception study and partly through close reading, inspired by the art-historical concept of queer abstraction. How is the portrayal of sexuality in Kropp handled in newspaper reviews 1980? How can the same depictions be interpreted through queer abstraction? By merging art historian David J. Getsy’s understanding of the concept of queer abstraction with art historian Alfred H. Barr’s older definitions of near-abstractions and pure-abstractions, the study shows how different abstract moves in Kropp operates. On the one hand, the near-abstractions camouflage lived queer experiences and desires, and on the other hand, the pure-abstractions present a more utopian queer sexuality, beyond heteronormative concepts. The study also shows how, in Eve Kosofsky Sedgwick’s sense, a more reparative than paranoid gaze at the review material reveals how queer-abstract understandings of the body and sexuality emerge in the reception, but without its sexual-political undertone. Additionally, the study shows how queer abstraction, paired with close reading, can function as a method for poetry research, particularly for literature that does not rely on prosaic functions such as narrative progressions and cohesive characters
Blickens väg i matematik : En eye-tracking studie om hur olika bildtyper påverkar problemlösning hos elever i årskurs 1
No full textBilder är en central del av matematikundervisningen, särskilt i de tidiga skolåren där visuellt stöd ofta används för att konkretisera abstrakta begrepp. Tidigare forskning visar att vissa bilder kan bidra till kognitiv överbelastning snarare än att stödja lärande. Syftet med denna studie är att undersöka hur elever i årskurs 1 tolkar och använder olika typer av bilder i matematiska problemlösningsuppgifter samt hur dessa påverkar deras kognitiva belastning. Studien har en kvalitativ ansats och bygger på data från eye-tracking, där 13 elever genomförde uppgifter med fyra olika bildtyper: symboliska, kontextuella, dekorativa och irrelevanta. Resultaten visar att symboliska och vissa kontextuella bilder stödjer problemlösning genom att minska split-attention-effekter och främja fokuserad växling mellan text och bild. Dekorativa och irrelevanta bilder ledde däremot ofta till splittrad uppmärksamhet, fler fixeringar och ökad extraneous load. Studien belyser vikten av noggrant bildval i läromedel och behovet av att undervisa elever i hur bilder kan användas som problemlösningsstöd
Bränna skrifter, forma nationer - Koranbrännandets påverkan på Sverige som nation
No full textThe purpose of this study was to analyze the main political ideas which are expressed through news-based media in Sweden. The study additionally tries to determine whether the groups within the study can be categorized and if they can, how? The studies material consists of five news articles from different media outlets which were studied to produce a result through idea analysis. The effect on how events are presented affects the receiver’s reaction to the news by the audience. Therefore, it is theoretically relevant to inquire how discourse is presented in the media. Based on the results of the idea analysis patterns of Quran burnings shows connection between portrayals of violence and connections to Muslim cultural markers used as examples of the security risk that right-wing populists used to portray Muslims. The categorization of groups was not possible due to separations within the specific groups based on political differences.