Blekinge Institute of Technology
Not a member yet
13576 research outputs found
Sort by
Voice as a Digital Biomarker : Machine Learning Applications for Chronic Obstructive Pulmonary Disease Assessment
No full textChronic Obstructive Pulmonary Disease (COPD) is a leading cause of morbidity and mortality worldwide, with high underdiagnosis rates due to limitations in current diagnostic methods such as spirometry. This doctoral thesis explores the potential of voice as a digital biomarker to support the assessment of COPD, guided by the principles of Applied Health Technology (AHT), which emphasizes interdisciplinary collaboration and real-world applicability. The research includes four interconnected studies. Study I presents a systematic literature review of machine learning (ML) applications for voice-affecting disorders, identifying COPD as underrepresented in current research. Study II addresses this gap by collecting a new dataset of vowel [a:] recordings from Swedish-speaking COPD patients and healthy controls once a week in self-determined quiet settings. Voice features, including baseline acoustic (BLA) parameters and Mel-Frequency Cepstral Coefficients (MFCCs), were extracted and used to train three ML classifiers: CatBoost (CB), Random Forest (RF), and Support Vector Machine (SVM). CB demonstrated the highest test accuracy at 78%. Study III investigates the effects of signal segmentation on model performance and shows that certain temporal segments of voice recordings contain more informative patterns, enhancing classification outcomes by increasing accuracy to 85%. Study IV applies statistical and practical significance tests to compare voice features between COPD and healthy groups. A total of 34 features, including shimmer measures and higher-order MFCC derivatives, were found to meaningfully differentiate the groups. This thesis reframes the human voice as a source of clinically relevant data, demonstrating how it can be digitized, analyzed, and interpreted using ML to aid COPD assessment. The results indicate that voice-based analysis can provide an accessible, non-invasive, and scalable complement to existing diagnostic tools. By integrating technical, clinical, and ethical perspectives, the thesis contributes new knowledge and practical methodologies that align with AHT's goal of creating value-driven, user-centered healthcare solutions. The findings support future development of mobile and remote voice-based screening tools for COPD and other conditions
A Transformer-Based Approach for Text Scam Detection with Synthetic Data Augmentation
No full textAs online communication becomes increasingly essential in everyday life, the risk of falling victim to digital scams has grown significantly. These scams are no longer limited to single deceptive messages; In reality, the number of digital scams is growing by the day. These scams are no longer solo deceptive one-time messages but rather involve a full conversation in which the manipulative spirit slowly exerts its influence on the users. In such cases, either the detection systems are not able to pick up contextual information or they are unable to follow ever-evolving patterns of language. The thesis addresses an AI-within approach beyond a single message exchange on which one can base their scam detection: synthetic data augmentation for better performance. Background: Online-scams-enabled types have now become more dynamic and cases harder to detect, especially when unfolding over multiple interactions. Rule-based and classical machine-learning-based systems do not cope well with the mutable nature and cues of conversational scams. Objectives: The objective of this thesis is to augment the aim of scamming detection systems with Transformer-based models for grasping the context of messages. By focusing on RoBERTa and examining whether synthetically generated scam conversations can serve as a data augmentation mechanism against rare real data to enhance model performance. Methods: A three-stage procedure was run: first, synthetic scam conversations were generated through language models to augment the training data. The Classifiers Machine learning and deep learning, including RoBERTa, were trained on this data; and third, model performance was measured on a mixture of real and synthetic samples. Results: Trained with synthetic augmentation, RoBERTa beat the classically used classifiers in recognizing the scam patterns that emerge in conversations. The results suggest that combining synthetic and real data helps the model to detect the subtle and evolving tactics of scam. Conclusions: This thesis demonstrates the use of context-aware models coupled with synthetic data as a scalable and effective approach towards identifying multiturn scams. This approach is a step toward creating safer digital communication systems and highlights the prospects synthetic augmentation has in low-data cybersecurity contexts
Green Investments and Firm Profitability: An Analysis of Nordic Energy Companies : A statistical analysis of sustainability investments in the Nordic region based on a linear regression evaluating share price, net profit as profitability measures
No full textDuring recent years, climate change has been a deeply investigated subject, leading various governments to ratify climate-related agreements, such as the Paris agreement. These agreements triggered an unprecedented transition towards sustainability in the energy industry, historically related to fossil fuels. The aim of this work is to understand the financial impact for the energy sector when investing within sustainability in the Nordics, as this region has been at the forefront when it comes to sustainability. Linear regressions were performed to evaluate correlations between firms’ financial indicators and potentially underlying variables such as oil price and green investments. Equinor, St1, Neste, Preem, and Shell were the firms considered in this study. First, the share of green investments was evaluated, identifying the different investment strategies that firms may adopt. In addition, it was identified that green investments have a strong correlation with the enforced environmental policies, with new policies correlated with an average of 5% increase of the share of green investments. The study continues evaluating the percentage of net profit, showing a positive correlation with the oil price, indicating that increased profits are correlated with higher oil prices. Equinor is an exception, as its green investments show a unique correlation with net profit, likely due to its unique portfolio of offshore wind power. Finally, the study considers the share price as an independent variable, which has a negative correlation with green investments, indicating a risk-averse position from the shareholders. Additionally, the share price has a positive correlation with environmental policies, which is in line with previous studies arguing that it is beneficial for the overall sector to have a more regulated market. Equinor is an exception also in this case, as it shows a positive correlation with green investments, probably due to the increased profitability
Hybrid microgrid systems for Energy Resilience : A Techno-Economic Assessment of a Fully Automated Industrial Logistics Centre
No full textThe transition to resilient and sustainable energy systems is a pressing priority for industrial actors, who face rising energy volatility, stringent climate policies, and tightly coupled supply chains. This thesis investigates whether hybrid microgrid configurations can improve the techno-economic performance and operational resilience of a fully automated logistics facility that operates continuously at full capacity. IKEA’s Malacky Packaging and Distribution Centre in Slovakia serves as the case study. The analysis is grounded in resilience engineering frameworks that distinguish resilience from reliability and conceptualise energy as a strategic supply chain asset. A quantitative, scenario-based methodology was employed, with simulations conducted in HOMER Pro v3.14 to compare a grid-only baseline with hybrid alternatives that integrate photovoltaic (PV), battery energy storage systems (BESS), and wind power. Results show that while the grid-only case offers the lowest near-term investment, hybrid microgrid systems deliver superior long-term outcomes. Scenario 1 (PV+BESS) reduces the Levelized Cost of Energy (LCOE) by approximately 25% compared to baseline while providing 15 hours of outage autonomy (Resilience Index, RI = 0.59). Scenario 2 (PV+BESS+Wind) achieves 16 hours of autonomy (RI = 0.71) and reduces LCOE by approximately 21% versus baseline but faces practical limitations due to local wind permitting and regulatory constraints. Based on these findings, Scenario 1 is recommended as the most viable configuration, striking a balance between substantial cost savings and improved resilience. In contrast, Scenario 2 remains a potential longer-term option under more favourable regulatory conditions. All Future research should advance methods such as stochastic outage modelling and intelligent energy management, while also testing applications across diverse industrial sectors
Advancing GNSS-RO Detection of Ionospheric Irregularities Using Refined Back Propagation and GOLD Data
No full textThis paper investigates on the detection and localization of ionospheric irregularities using GNSS Radio Occultation (GNSS-RO). We propose a new segmented phase screen (PS) approach to improve vertical and horizontal localization and remove the presence of outliers. The study focused on the May 2024 geomagnetic solar storm is presented, consisting of a comparison of the GNSS-RO back propagation (BP) irregularity positioning against the data of NASA’s Globalscale Observations of the Limb and Disk (GOLD) mission. This study is performed for validation purposes and examines the presence of equatorial plasma bubbles (EPBs) at predicted locations. Experimental RO data from EUMETSAT’s MetOp satellites is used to demonstrate the method’s capability to characterize the distribution of ionospheric irregularities. Results validate the segmented approach's capabilities of detecting irregularity structures and identifying their centroids with improved performance compared with the previous version of the algorithm.
An evidence-based neuro-symbolic framework for ambiguous image scene classification
No full textIn this study, we propose a novel neuro-symbolic approach to deal with the inherent ambiguity in image scene classification, combining the usage of pre-trained deep learning (DL) models with concepts from modal logic and evidence theory. The DL models are used to detect objects and estimate their depth in a set of labeled images. The obtained outputs are employed to form a dataset of instances characterizing the possible classes. Subsequently, a multi-valued mapping is defined between the data instances and the considered images resulting into each image being represented by the set of instances associated with it. The obtained mapping is utilized to infer necessity and possibility conditions of each class, or equivalently its upper (plausibility) and lower (belief) probabilities. Based on these interval evaluations, a rule-based and a score-based classifiers are built. The overall method is explainable and directly interpretable, robust to data scarcity and data imbalance. The presented framework is studied and evaluated on an abandoned bag detection use case.
Staden och det gröna : En studie om hållbar stadsutveckling och användning av ekosystemtjänster i förtätade stadsmiljöer - en fallstudie på Karlskronas innerstad.
No full textUnder de senaste 30 åren har arbetet med att skapa grönare och mer hållbara städer varit i fokus, där ekosystemtjänster - naturens bidrag till människors vällfärd - haft en central roll i planeringen. Grönområden i städer är viktigt för både folkhälsa och social jämnlikhet, särskilt för socioekonomiskt utsatta grupper. I takt med att urbaniseringen ökar, förväntas 70% av världens befolkning bo i städer år 2050, vilket ställer höga krav på hållbar stadsplanering. Grönstrukturen i städer bidrar med viktiga funktioner som biologisk mångfald, klimatreglering, dagvattenhantering och sociala mötesplatser. För att optimera dess roll i framtida stadsutveckling krävs mer kunskap om hur den gröna infrastrukturen kan användas som ett verktyg för att skapa socialt och ekologiskt hållbara stadsmiljöer. Studien syftar där av till att undersöka hur den gröna infrastrukturen och ekosystemtjänsterna kan intergreras på förtätade stadsmiljöer. Arbete tar upp ett fyra designprinciper som hjälper att underlätta för planerare att gestalta miljöer som uppmuntar för användningen av gröninfrastruktur och ekosystemtjänster. Slutsatsen av denna studie är att den gröna infrastrukturen och ekosystemtjänsterna behöver få en större roll i den nuvarande stadsplaneringen, detta för att skapa mer levande och hållbara stadsmiljöer då urbaniseringen kommer fortsätta att öka
Teachers’ Strategies and Technology Use for Enhancing Students’ Critical Thinking in Nursing Simulation-Based Learning : A Qualitative Pilot Study
No full textIn nursing education, simulation-based learning (SBL) is often used to bridge theoretical knowledge and practical application, supporting nursing students in developing their critical thinking (CT) skills. Despite the benefits of using SBL in nursing education, research gaps remain in understanding student learning outcomes. Furthermore, there is a lack of studies describing the specific use of technology in its application. The objective of this study was to explore strategies for learning and the use of technology to enhance nursing students’ CT within the SBL context. This research was conducted as a qualitative pilot study, using a semistructured interview technique to gather insights from teachers at 2 universities in the south of Sweden. The obtained data were analysed in accordance with the phenomenographic analysis introduced by Sjöström and Dahlgren. The results revealed participants’ perceptions of useful strategies for student learning and different ways of using technology. In particular, the results are reflected in 5 descriptions of categories: motivating environment, facilitating preparations, active participation, student-centeredness and reflective observations. While the findings may not be directly applicable to clinical practice, the study’s findings offer examples of effective strategies for student learning and technology use, thus providing valuable guidance for teachers implementing SBL in nursing education. To gain a more comprehensive understanding of using SBL as a teaching method, future research should aim to investigate nursing students’ experiences of how CT is promoted via simulations.
Key success factors for commercializing biotechnological innovation : a post-COVID analysis of biopharma start-ups in Sweden and Germany
No full textThe European biotechnology sector plays a pivotal role in advancing healthcare innovation, with biopharma start-ups serving as key agents in translating scientific research into commercial medical solutions. Yet, numerous start-ups face persistent barriers to commercialization, ranging from limited access to funding and long regulatory cycles, to the challenge of building viable business models. These hurdles were further exacerbated by the COVID-19 pandemic, which reshaped global investment patterns, disrupted clinical pipelines, and triggered regulatory adaptations across Europe. This thesis investigates the key success factors that enable biopharma start-ups to navigate commercialization challenges, particularly in the context of the pandemic and its aftermath. It focuses on Sweden and Germany as case studies of the broader European biotech landscape due to their established innovation ecosystems, strong public-private research networks, and dynamic investment environments. Special attention is paid to two defined timeframes: the COVID-influenced period (January 2020–December 2021) and the post-pandemic period(January 2023–December 2024). A mixed-methods approach was adopted in which qualitative insights obtained from semi-structured interviews, explored perspectives on how start-ups responded to regulatory and market disruptions, funding gaps, and strategic uncertainties. These were complimented by quantitative data from an online survey targeting a broad range of stakeholders; founders, investors, incubators, policymakers, and business development executives from large pharmaceutical companies involved in start-up partnerships or acquisitions. Findings reveal that successful biopharma start-ups actively leveraged strategic partnerships, engaged early with regulators, diversified their funding sources, and aligned their commercialization models with shifting policy and market conditions. Government interventions such as recovery-oriented investment schemes, regulatory fast-tracking, and start-up incubation support were found to be instrumental in influencing success trajectories. In contrast, start-ups lacking proactive business strategies or adaptive capacity struggled to attract investment or gain regulatory traction. Importantly, this study identifies distinct commercialization patterns between the two periods: during the pandemic (2020–2021), start-ups prioritized regulatory agility, pandemic-related pivots, and emergency funding opportunities, while in the post-pandemic period (2023–2024), emphasis shifted toward strategic scalability, broader market positioning, and increased challenges in attracting sustainable funding. This study contributes to the broader understanding of biotech commercialization in Europe by identifying core business, regulatory, and financial factors that support start-up viability in dynamic, high-risk environments. Its findings provide actionable recommendations for biotech entrepreneurs, investors, and policymakers to strengthen start-up resilience, improve market readiness, and foster innovation-driven growth across the European biotech sector
Thermal shock effect on microstructure, electrical properties and corrosion performance of ultrasonically welded wire harness-terminal joints
No full textWith the rapid development of new energy vehicles and power electronics industry, the stability and durability of automotive wiring harness connections have become key technical issues. In this paper, a thermal shock test is designed for aging phenomenon caused by extreme temperature conditions of 25 square copper wire harness-terminal connectors in actual use of vehicles. The results show that after 72 thermal shock cycles, the grain organization of copper wire inside connector is significantly refined and the percentage of large-angle grain boundaries is increased. In addition, the distribution of cubic texture orientation {100} <001> weakens and transforms towards the shear texture {111} <111> parallel to the Z direction of common grain organization. In tensile test, the average tensile strength of joints after thermal shock increased by 6.08 %. The resistance of aged joints increased by 14.75 %, but still met the electrical safety standards. The temperature rise of joint resistance increased by 3.81 degrees C after being connected to a constant current of 100A and lasting for 2.5 h. In terms of corrosion resistance, the improvement of joints after thermal shock test is related to microstructural changes in joint material. Grain refinement introduced a high fraction of grain boundaries and more grain boundaries were able to act as barriers to prevent the penetration and diffusion of corrosive media