Publikationer från Linköpings universitet
Not a member yet
58266 research outputs found
Sort by
Förenkling av omvandling från strukturerade rapport-mallar i PACS till openEHR-baserade rapporter
No full textInteroperability within the healthcare sector means the ability to share medical data in an organized manner. To achieve interoperability there is a need for structures and standards on multiple levels, including the semantic level. openEHR is an open framework that works to improve this area by providing technical standards for data within an electronic health record, but these standards are not yet widely used in Sweden. There is, however, already a lot of medical data being produced. One method used for writing reports within the healthcare sector is Structured Reporting (SR). These reports get their structure from SR templates and the fields in these templates can in theory be matched to fields in an openEHR template to make the report openEHR compatible. The process of matching fields between SR templates and openEHR templates is known as mapping. This thesis project investigates what the mapping process looks like today, what difficulties are associated with it, and how they can be alleviated. After interviewing informaticians it was found that the work they do is still largely manual in nature and could benefit from partial automatization. When two fields are being mapped, the person doing the mapping has to write out or copy the path to each field. This can be a time consuming task, especially for more complex templates where a deep structure creates long paths. From this it was decided the thesis would develop a tool with a graphical user interface which can produce and store the correct paths for mapped fields. Another issue with the state of mapping is the differences between SR and openEHR. A secondary focus of this thesis is thus looking at to what extent mapping is possible and why any unmapped fields could not be mapped. The resulting tool is a web application that can produce a mapping configuration file based on user input. The application takes SR and openEHR templates as input and after pre-processing, displays them in two separate tree structures. The user can then click on a node in each tree and map these to each other. The main benefits of the web application are the simple graphical user interface, the automatic generation of paths, and the pre-processing of the SR templates. To further reduce the manual labor associated with the mapping process future efforts should be made on implementing artificial intelligence and natural language processing. However, it should be stated that the real goal is for medical data to be openEHR compatible to begin with, completely erasing the need for mapping in the first place. The mappings possible in the application did not achieve 100% coverage for any of the SR or openEHR templates used in this project. While some of the disconnect can be erased by implementing further functionalities in the application, the results still show that there is currently a disconnect between the two structures. Neither SR nor openEHR has a widespread use, and as more work is done on implementing both SR and openEHR in healthcare there should be an awareness of the other structure to reduce the disconnect at the source. The effect that SR and openEHR have on interoperability today is therefore hard to conclude
Optimizing surveillance missions for UAV fleets by column generation
No full textThis thesis explores the use of column generation methods to solve an extended vehicle routing problem tailored to surveillance missions using UAVs. The goal is to minimise the number of UAVs required by finding optimal routes while still meeting surveillance requirements. A mission involves a fixed set of locations that must be continuously monitored and a number of depots from which UAVs can be deployed. The problem takes into account factors such as location-specific time constraints, a heterogeneous UAV-fleet and multiple depots. Two classes of problem instances are studied. In the first, multiple UAVs are allowed to collaborate on a single tour to satisfy strict time constraint on revisits. In the second, a single UAV may perform multiple tours, where the revisit time requirements are less strict. The goal is to solve these problems using column generation and demonstrate their practical value in real-world applications. To evaluate the models, fictional instances with a varying number of depots and monitoring locations were constructed. An upper time limit was used for all computations. The results show that the models are capable of producing useful solutions. In particular, the second class of problems proved significantly easier to solve within the given time frame. However, for larger problem instances, timeouts were frequent due to the computation time required to solve large subproblems. Lastly, a post-processing step was used on all solutions to reduce total route times and thus improve usefulness in real-world applications.Denna uppsats utforskar användningen av kolumngenereringsmetoder för att lösa ett utökat fordonruttningsproblem anpassat för övervakningsuppdrag som använder UAV:er. Målet är att minimera antal UAV:er som krävs genom att hitta optimala rutter som uppfyller övervakningskraven. Ett uppdrag består av ett fast antal övervakningsplatser som måste övervakas kontinuerligt samt ett antal depåer från vilka UAV:erna kan starta sina uppdrag. Problemet tar hänsyn till faktorer såsom platsberoende krav på återbesökstider, en heterogen UAV-flotta och tillgång till flera depåer. Två typer av problemklasser studeras. I den första tillåts flera UAV:er samarbeta på en gemensam tur för att uppfylla strikta tidskrav för återbesök. I den andra kan en enskild UAV genomföra flera rutter, där kraven på återbesök är mindre strikta. Målet är att lösa dessa problem med kolumngenerering och påvisa deras praktiska relevans i verkliga tillämpningar. För att utvärdera modellerna konstruerades fiktiva instanser med varierande antal depåer och övervakningsplatser. En tidsgräns tillämpades på alla beräkningar. Resultaten visar att modellerna kan generera användbara lösningar, men för större instanser förekom dock frekventa tidsavbrott, främst på grund av den höga beräkningstiden vid lösning av stora delproblem. Det noterades att den andra problemklassen var avsevärt lättare att lösa inom den givna tidsramen. Slutligen användes ett efterbearbetningssteg på samtliga lösningar, för att minska de totala ruttiderna samt för att öka lösningarnas användbarhet i praktiken
Student Assistants in Schools for Students with Intellectual : Disabilities: Roles, Collaboration, and Educational Needs
No full textInom skolformen anpassad grundskola arbetar förutom lärare även elevassistenter. Yrkesgruppen elevassistenter har ett komplext uppdrag som omfattar både omsorg och stöd i lärandet. De samarbetar med lärare och andra elevassistenter i arbetslag. Trots det ansvar och komplexa uppdrag elevassistenter har, saknar de ofta pedagogisk utbildning och nämns inte i skolans styrdokument. Det råder ett bristande forskningsläge vad gäller yrkesgruppen, särskilt inom anpassad skola. Studiens syfte var att undersöka hur elevassistenterna uppfattar sin roll och identifiera framgångsfaktorer och hinder i samarbete med läraren, samt att undersöka eventuella behov av utbildning och kompetensutveckling. En kvalitativ metod användes och materialet samlades in genom semistrukturerade intervjuer. Nio intervjuer har genomförts med erfarna elevassistenter inom anpassad skola. Det insamlade materialet analyserades tematiskt med ett professionsteoretiskt perspektiv. Resultaten visar att goda relationer har betydelse för trivsel och samarbete i arbetslaget samt vikten av en tydlig organisation som möjliggör detta samarbete. Vår studie bekräftar elevassistenternas komplexa uppdrag och vi ser att detta uppdrag kräver många kompetenser. Dessa kompetenser kräver kunskaper som antingen erhålls genom utbildning eller kompetensutveckling, något som ofta saknas. Vidare visar resultatet att elevassistenterna önskar tydliggöra sitt uppdrag samtidigt som de önskar ökad delaktighet och ett utökat ansvar. Studiens resultat visar att elevassistenterna vill synliggöras och erkännas av rektor och andra beslutsfattare. De efterfrågar en ökad status. Vi tolkar dessa resultat som att elevassistentyrket behöver genomgå en professionaliseringsprocess. Vidare forskning på området föreslås eftersom elevassistenter i anpassad skola är ett outforskat område, men också för att ökade kunskaper kan bidra till en utveckling av elevassistentyrket. Detta kan leda till en kvalitetshöjning för eleverna i den anpassade skolan
Transition from Preschool to Compulsory School or Compulsory School for Children with Intellectual Disabilities : A Comparative Study of two Municipalities in Sweden
No full textSyftet med denna studie var att studera övergången mellan förskola och skola för barn med intellektuell funktionsnedsättning. Studien har fokuserat på hur övergången går till i två olika kommuner i Sverige, vad som är betydelsefullt och vad som utmanar i processen enligt de yrkeskategorier som deltar, till exempel förskollärare, lärare, rektorer och specialpedagoger. Studien har en kvalitativ ansats och har genomförts med hjälp av semistrukturerade intervjuer. Dokument som har med övergången att göra har också samlats in. Analysstrategin som användes var tematisk analys. Bronfenbrenners systemteori utgör studiens teoretiska ramverk. Resultatet visar att de båda kommunerna har upprättat en organisation för hur övergången går till. Organisationen ser olika ut. En skillnad är vilka aktörer som är delaktiga. Vad som tydligt framkommer för att övergången ska bli lyckad är samverkan med hemmet och externa aktörer, att göra vårdnadshavare trygga i processen samt att genomföra individanpassade övergångsaktiviteter både för barn och vårdnadshavare. Resultatet visar även att det finns dilemman och utmaningar i övergången såsom organisatoriska faktorer för att genomföra övergångsaktiviteterna, samt att ”hitta lämplig väg” in i skolan för barn med intellektuell funktionsnedsättning, då förskoleklass ej finns i anpassad grundskola. Studiens slutsats är att övergångar för barn med intellektuell funktionsnedsättning upplevs få goda förutsättningar att bli framgångsrika när det finns en tydligt strukturerad organisation kring övergången, både gällande ansvar och utifrån tidsperspektivet. Framgångsfaktorer och en implikation till specialpedagoger som arbetar med övergångar är: betydelsen av en tydlig organisation i övergångsarbetet, att samverka, att starta tidigt, att skapa trygghet samt att använda övergångsaktiviteter formade utifrån barnets och vårdnadshavares behov
Outdoor Education’s Integration in Teacher Formation: a Range of Perspectives : A phenomenographic study on outdoor education practitioners’ perspectives on the integration of outdoor education in teacher formation
No full textThis study aimed to examine the range of Outdoor Education (OE) practitioners’ perspectives on how OE should be integrated in Teacher Formation (TF). This research follows a phenomenographic approach in which OE practitioners from various countries throughout Europe were interviewed in order to determine how they perceived the possible integration of OE in TF. An outcome space was used to represent the findings of the research, allowing correlations and differences amongst the various perspectives to be brought to light. The findings revealed a spectrum of viewpoints, influenced by the practitioners’ own foundational views on OE. Key themes include the importance of embedding OE either across subjects or as a stand-alone course, the relevance of both theory and practice in training, and the significance of place in outdoor learning environments. Additionally, the study identified a range of practical skills considered essential for preservice teachers, such as organizational and collaborative skills, adaptability, creativity, safety awareness, and confidence in nature-based settings. Challenges related to time, resources and curriculum constraints as well as cultural attitudes were also discussed, along with recommendations such as institutional support and educator networks
TikTok: The new political arena for young voters : TikTok as a political marketing platform for building brand awareness among young voters for swedish political parties
No full textI takt med att TikTok blivit en central kanal för informationssökning bland unga väljare har det svenska politiska landskapet förflyttat sig till en plattform där snabba, underhållande och visuella element prioriteras. Studien syftar till att undersöka hur svenska partier, Socialdemokraterna och Moderaterna, använder TikTok för att bygga varumärkesmedvetenhet bland unga väljare i åldern 18-25år. Studien tar avstamp i teorier om politiskt varumärke, mänskligt varumärke och marknadsföring i sociala medier (MSM) med fokus på fyra dimensioner: Underhållning, trendighet, interaktion och elektronisk mun-till-mun. Studien har en deduktiv ansats med en kvalitativ metod där semiotiska observationer av partiernas TikTok-innehåll, och semistrukturerade intervjuer med unga väljare utgör datainsamlingen. Sammanställningen av data genomfördes via öppen kodning, innehållsanalys och tematisk analys. Slutsatsen för studien är att svenska partier använder TikTok som ett verktyg för att bygga varumärkesmedvetenhet hos unga väljare genom att kombinera ideologiska budskap, underhållande inslag och partiledarens personlighet. Genom att anpassa innehållet till plattformens premisser skapas igenkänning, förtroende och engagemang hos unga väljare. Rekommendationerna från studien är att svenska partier bör kombinera värderingsdrivna budskap med tydliga visuella element och lyfta fram partiledaren såväl som professionella som relaterbara. Partierna bör använda humor och trender på ett autentiskt sätt och interagera med användare på TikTok. Allt för att bygga en enhetlig, trovärdig och engagerande varumärkesidentitet som bidrar till partiets varumärkesmedvetenhet.As TikTok has become a central platform for information-seeking among young voters, the Swedish political landscape has shifted to a space where fast-paced, entertaining, and visual content is prioritized. This study aims to examine how the Swedish political parties, the Social Democratic Party and the Moderate Party, use TikTok to build brand awareness among young voters aged 18–25. The study is grounded in theories of political branding, human branding, and social media marketing (SMM), with a focus on four key dimensions: entertainment, trendiness, interaction, and electronic word-of-mouth. Adopting a deductive approach and a qualitative methodology, data was collected through semiotic observations of the parties’ TikTok content and semi-structured interviews with young voters. The data was analyzed using open coding, content analysis, and thematic analysis. The study concludes that Swedish parties use TikTok as a tool to build brand awareness among young voters by combining ideological messages, entertaining elements, and the personality of the party leader. By adapting content to the platform’s logic, parties generate recognition, trust, and engagement among young audiences. The study recommends that Swedish parties combine value-driven messages with clear visual elements, and present their party leaders as both professional and relatable. They should use humor and trends authentically and engage actively with users on TikTok to build a coherent, credible, and engaging brand identity that strengthens party brand awareness
Fusion of Radiographic Images and Electronic Health Records for Classification of Femoral Fractures using Transformers
No full textOsteoporosis is a geriatric disease characterized by decreased bone density, commonly treated with bisphosphonates. This class of drugs inhibits bone resorption and increases bone mineral density, effectively reducing overall fracture risk. However, long-term bisphosphonate therapy has been associated with atypical femoral fractures, which are insufficiency fractures that can occur with minimal or no trauma. This effect necessitates careful assessment when considering extended bisphosphonate treatment, and therefore accurately classifying such fractures from normal femoral fractures that usually stem from high-trauma impact is of clinical concern. This thesis explores deep learning models, namely Transformer-based models, to aid in classifying atypical femoral fractures from normal femoral fractures. Fusion of two modalities, radiographic images of fractures and electronic health records of the patients, was performed to allow the networks to make predictions using more available information. For this, a vision transformer and a tabular transformer models were employed. The fusion was done using a conventional fusion strategy, but also an attention-based one. The dataset used comprises data from 1, 073 patients in Sweden who suffered a femur fracture, with radiographs obtained from 72 hospitals in 2011, totaling 4, 014 images. All images have a fracture present. This was coupled with detailed register information from the Swedish National Patient Register. The data was preprocessed using common techniques for each respective modality and split on patient-level using 5-fold cross-validation. Five models were employed to perform the binary classification of the fracture, (1) unimodal vision transformer, (2) unimodal tabular transformer, (3) multimodal conventional late fusion, (4) multimodal intra self-attention fusion, (5) multimodal inter cross-attention fusion. The models were assessed using performance metrics (Accuracy, AUC, F1-score, Matthews Correlation Coefficient) and prediction uncertainties from MC dropout. Model predictions were also aggregated from the image-level to the patient-level using the inverse uncertainties to increase clinical relevance of the results. Model evaluation was performed using the DeLong test to compare AUCs and Wilcoxon rank-sum test to compare uncertainties. Among the models, the unimodal vision model had majority of best test metrics at the patient-level when averaged over the folds, with an accuracy of 99.07%, F1-score of 0.9714, and an MCC of 0.9665. The multimodal conventional late fusion model and the multimodal inter cross-attention fusion model had the highest average AUC of 0.9994. There were no significant differences between the AUC of the respective models. The comparison of uncertainty showed significant outcomes for all tests except for the conventional late fusion compared to the multimodal inter cross-attention fusion model. In conclusion, the unimodal vision model performed best according to the performance metrics, however when accounting for uncertainties, the fusion models showed an increased performance.
Dynamisk Identitets-Baserad Routing i SDN-Kontrollerade VPLS
No full textThe increasing demands for mobility, scalability, and security in industrial networks challenge the limitations of traditional Virtual Private LAN Services (VPLS), which rely on static MAC-based forwarding. This thesis explores the possibility to integrate the Host Identity Protocol (HIP) with Software-Defined Networking (SDN) to enable dynamic identity-based routing in SDN-controlled VPLS environments. A prototype controller was implemented using Ryu and evaluated in a simulated network based on real-world topologies. Two routing policies—shortest path and least hop—were compared using latency and hop count as performance metrics. Results show that dynamic routing based on Host Identity Tags (HITs) enables per-connection policy selection, allowing for trade-offs between latency and privacy. While shortest path routing offered better performance, least hop routing reduced exposure to intermediary nodes, enhancing privacy. The findings suggest that identity-based routing in SDN-controlled VPLS can offer flexible, context-aware routing decisions suitable for industrial applications requiring both security and efficiency
Utvärdering av användningen av GPT-4o för att migrera ostrukturerad adressdata till ISO 20022-kompatibilitet
No full textThis thesis investigates the potential of GPT-4o, a large language model, in converting unstructured postal address data into ISO 20022-compliant structured formats. ISO 20022 is a globally recognized standard crucial for financial messaging interoperability and security. The research explores the challenges of standardizing diverse, unstructured postal address formats and evaluates GPT-4o's performance using metrics like precision, recall, and F1-score. Two datasets were created using DeepParse address data, one dataset for evaluating initial performance and identifying problems to further refine the prompt. The other dataset was used together with the refined prompt for final evaluation. The findings reveal that GPT-4o demonstrates high accuracy in structuring address data, achieving an average F1-score of over 96%. However, challenges remain in handling numerical elements and country-specific formats. Refinements in prompts and country-specific examples show promise for improving accuracy. This study highlights GPT-4o's potential for address parsing, offering a potential solution for organizations transitioning to ISO 20022 compliance. Future work will focus on refining datasets for more accurate results and exploring tailored prompts to enhance performance further
Utvärdering av Starlinks prestanda : med hinder och i rörelse
No full textThis study evaluates the performance of Starlink satellite internet service under conditions of obstruction and motion. With the increasing reliance on stable and high speed internet connectivity for various applications, understanding the limitations and capabilities of emerging technologies like Starlink is critical. This research aims to provide insights into how obstructions as well as the movement of the Starlink dish, affect the overall performance and reliability of the Starlink network. By conducting experiments using a Starlink kit, we assess the performance of Starlink under different conditions: without obstruction, with obstruction, and in motion. The tests were carried out by setting up the Starlink equipment in various locations and also by moving it using a trolley to simulate motion. We utilized tools such as iPerf3 and Wireshark to capture and analyze network performance data, focusing on metrics like bandwidth and retransmissions. The results indicate that obstructions significantly impact Starlink’s performance, causing severe throughput fluctuations, increased TCP retransmissions and a different pattern in the distribution of these retransmissions. In motion, Starlink experiences throughput drops that recover over time, though with slightly reduced overall performance compared to the baseline. The study suggests that obstruction and motion introduce considerable variability in connection stability. Future work should explore extended test durations, higher motion speeds, and varied locations to comprehensively evaluate Starlink’s performance under different conditions