36270 research outputs found
Sort by
Sensor Fusion based Health Management Systems for Semiconductor Power Modules in Electric Vehicle Inverters
No full textPower semiconductor devices in electric vehicle inverters face critical reliability challenges, with temperature-related aging issues accounting for 55% of converter failures. Accurate junction temperature monitoring is essential for managing aging mechanisms and preventing catastrophic failures. Insofar, existing single Temperature Sensitive Electrical Parameter (TSEP) approaches suffer from the limitation due to high measurement noise susceptibility, sensitivity to operating conditions, and inability to decouple temperature effects from device degradation. With such situation in mind, this thesis develops a novel multi-TSEP fusion framework for SiC power modules for the purpose of systematically analysing how working current, voltage, gate parameters, and wear-out conditions have effects on TSEP characteristics.Two methods based on the Principal Component Analysis-Multiple Linear Regression (PCA-MLR) are proposed to address aforementioned limitations. The first method tackles the noise susceptibility problem by extracting and fusing 16 statistical features from single gate current waveforms, using PCA to eliminate multicollinearity and enhance robustness against measurement disturbances. The second method addresses both operating condition sensitivity and temperature-degradation coupling by integrating multiple TSEPs across different measurement channels, leveraging complementary temperature-dependent characteristics from various electrical parameters such as on-state voltage drop, gate threshold voltage, and switching transients to achieve condition-independent estimation.Extensive validation demonstrates that both methods maintain temperature estimation errors within 4-5°C even under severe noise conditions with input errors up to 40-50°C—an order lower of magnitude improvement than conventional single-TSEP approaches. The multi-channel fusion approach particularly excels in decoupling temperature effects from device aging, enabling simultaneous health monitoring and thermal management. These results establish a practical pathway for robust temperature monitoring in the power device converter system for next generation, particularly for electric vehicle applications where reliability is paramount.</p
Localising Western Transnational Higher Education in the GCC: Instructor Agency, Third Space, and Culturally Relevant and Responsive Teaching and Learning
No full textThis study explores the localisation of [w]estern transnational higher education (TNHE) in the Gulf Cooperation Council (GCC) region, focusing on expatriate instructor agency, Third Space theory, and culturally relevant and responsive pedagogies (Gay, 2002; Ladson-Billings, 1995).Grounded in critical theory, it examines how expatriate instructors adapt [w]estern business curricula to reflect the sociocultural realities of learners at an international branch campus in Qatar, framing the TNHE classroom as a Third Space (Bhabha, 2012), a site of hybridity, negotiation, and mutual cultural exchange.A qualitative case study combined a modified lesson study (adapted from Dudley, 2014) with structured autoethnographic vignettes (adapted from Pitard, 2016) to examine curriculum localisation and reflect on instructor identity, agency, and practice in TNHE. The study also drew on the Behavioural Engagement Related to Instruction protocol (Lane and Harris, 2015) and Patel’s (2017) Principles of Global Engagement to assess pedagogical adaptation, student engagement, and the presence of Third Space.Data collected through questionnaires, interviews, lesson artefacts, classroom observations, and instructor and student feedback were thematically analysed. In addition, the Principles of Global Engagement served a reflexive function in combination with the vignettes, illuminating instructor identity, institutional positioning, and curriculum adaptation.Findings indicate that, despite significant institutional constraints, expatriate instructors can foster inclusive, contextually relevant teaching practices. Student engagement improved when lessons incorporated culturally familiar content and interactive strategies, though sustained scaffolding is necessary to reduce reliance on rote memorisation.Building on Jabbar and Hardaker’s (2013) Five Pillars framework, this research contributes a Transnational Teaching and Learning model for implementing a culturally responsive approach to teaching and learning in TNHE. It offers original insights into TNHE scholarship by focusing on under-researched business classrooms, illustrating how instructor reflection and agency can transform [w]estern curricula into more equitable and locally meaningful education.</p
TriFTM-Net: Tri-Path Fourier-Temporal Modulation Network for macular edema pathology segmentation and reconstruction in high-precision intraoperative navigation
No full textOphthalmic diseases such significantly impair the vision of numerous individuals globally. Accurate and real-time 3D reconstruction of macular edema and retinal tears is crucial for improving surgical efficiency and success rates. However, lesion areas often exhibit considerable noise and high heterogeneity, and the imaging devices employed may introduce electronic noise and artifacts. Current 2D medical image segmentation techniques fail to achieve optimal outcomes. To overcome these challenges, we propose the Tri-Path Fourier-Temporal Modulation Network (TriFTM-Net). TriFTM-Net synergistically integrates spatial, frequency, and spatiotemporal features. This design effectively augments both feature representation and extraction. TriFTM-Net comprises three critical modules: the Tri-Path Spectral Hierarchical Encoder (TPSHE), which amplifies feature representation by integrating tri-path features; the Feature Re-Modulation (FRM), which reduces noise interference and enhances feature extraction; and the Hierarchical Feature Reconstruction Module (HFRM), which improves detail preservation in upsampled images. Comparative analysis with thirteen baseline methods demonstrates that our approach achieves the highest Dice scores, IoU, and Kappa coefficient on the OIMHS dataset.Our code is publicly available at https://github.com/IMOP-lab/TriFTM-Net.</p
Development of a blood test for uterine sarcoma-Diagnosis and monitoring (DOORS-D and DOORS-M) studies.
No full textUterine sarcomas can be difficult to differentiate from uterine fibroids due to many shared symptoms and imaging features, which can result in delayed or missed diagnosis, or over treatment. The 'Development of a blood test for uterine Sarcoma - Diagnosis' (DOORS-D) (ISRCTN14800787) and 'Development of a blood test for uterine Sarcoma - Monitoring' (DOORS-M) (ISRCTN14174468) studies aim to explore the role for circulating tumour DNA (ctDNA) to diagnose and to monitor uterine sarcomas. DOORS-D will recruit patients who have a suspected uterine sarcoma or large fibroid and are due to undergo surgery (hysterectomy or myomectomy) for a blood sample prior to surgery whereas DOORS-M will recruit patients who have been diagnosed with a uterine sarcoma in the previous 10 years for longitudinal blood sampling every 3-6 months over the course of the study. Information will be collated on patient characteristics and symptoms, tumour characteristics and diagnostic imaging, with representative images selected and analysed using large language models. Analysis of genomic/methylation profile of ctDNA samples collected from DOORS-M will be used to design a ctDNA-based 'test'. Analysis of the samples collected from the participants recruited to the DOORS-D study will enable the accuracy of the 'test' to differentiate uterine sarcomas from fibroids to be determined. In addition, the opinions of patients with a suspected or confirmed sarcoma will be explored through semi-structured qualitative interviews. Purposeful recruitment strategy will ensure that the experiences of women from diverse socioeconomic, cultural and ethnic backgrounds are included. The results of the studies will be shared through conference presentations and peer-reviewed publications.</p
Digital Dust: Tales of the Unexpected
No full text"Digital dust: Tales of the Unexpected" is a video ‘story’ in which I showcase waste from my (virtual) cutting room floor to reveal hidden narratives of archaeological practice. This is a creative response from a year of significant laser scan data processing. As I worked, I reflected on the large amounts of digital data we generate but never use, and rapidly changing software and formats. I paired the scan data visuals with Jean Michel Jarre’s laser harp music from the album Equinoxe to evoke the sense of magic I felt analysing the data and the transcendence of time and space. I was also keen to acknowledge importance of harps in storytelling. My images are simply screen grabs of things I found interesting as I worked, and the captions are snippets of thought, memory, and amusement sparked during the process. Together, I hope they form a playful reflection of light, time, beauty, and the complexity of archaeology in the digital age. This video was produced for the "Narrate Create" session of the "Theoretical Archaeology Group Conference 2024" (TAG 2024) held in Bournemouth University. Laser scan data were collected using a variety of Leica laser scanners, and analysed using Leica and Microsoft software under licence to the University of Leicester. The video forms one of a group of alternative archaeological stories and associated popular articles from TAG 2024 session participants ("Narrate Create" and "Stories as Old as Time") published in British Archaeology 2026 (April onwards). The overview article for the first group of stories is entitled "Narrate-create: stories as old as time" by Laura Basell, Kirsty Lilley, Fiona Coward, Lusia Zaleskaya, and Neil Redfern. The popular article which accompanies and explains this video by Laura Basell is entitled "Digital Dust? Finding unexpected stories in laser scans".Acknowledgements: Basell gratefully acknowledges the support of the British Academy through a mid-career fellowship (Reference: MFSS24\240089), which provided the time and space to reflect on the role of storytelling and narrative in archaeological contexts. TAG 2024 organising committee and the participants in our sessions for contributing to an engaging and thought-provoking conference.Fieldwork Collabortors Abdallah Khamis, Zanzibar Heritage Foundation and Dr Lee Bray, Dartmoor National Park Authority</p
Forecasting Inflation in the Presence of Structural Breaks: A Time-Varying Parameter Approach
No full textThis paper extends the work of Hall et al. (2025), who demonstrated that, while structural break tests for an unknown break date are unable to detect structural breaks near the end of a sample period, they can be effective when the user has a prior expectation of where the break occurs. In this paper, we demonstrate that a time-varying parameter forecasting model can also be effective when we know approximately where the break occurs. We use a Kalman filter time-varying AR forecasting rule, where the degree of time variation is governed by the Q-matrix. We provide evidence that this is a better formulation than the standard rolling window approach in the literature.</p
Dynamic and Ongoing De Novo L1 Retrotransposition Contributes to Genome Plasticity and Intrapatient Heterogeneity in Ovarian Cancer
No full textLong interspersed element-1 (L1) retrotransposons are the only protein-coding active transposable elements in the human genome. Although typically silenced in normal cells, they are highly expressed in many human epithelial cancers, including high-grade serous ovarian cancer (HGSC), and can integrate into the genome through retrotransposition. De novo L1 insertions are known to contribute to genomic instability and cancer evolution in epithelial malignancies, including HGSC, suggesting that they might also play a role in intrapatient tumor heterogeneity. In this study, we quantified de novo L1 insertions in clinical HGSC specimens and uncovered high heterogeneity in total L1 insertion events (L1 burden) between patients. HGSC tumors with high L1 burden were highly proliferative, whereas tumors with low or no L1 insertions showed enrichment of immune response and cell death pathways. Although the overall L1 burden was similar across different tumor sites within the same patient, the specific L1 insertions (L1 profiles) diverged significantly more than their single-nucleotide variants profiles. Taken together, these findings demonstrate that L1 activity and retrotransposition are highly dynamic in vivo and can contribute substantially to tumor genome plasticity, especially at late stages of cancer progression. The patient-specific propensity of acquiring L1 insertions (L1 burden) could be driven by molecular properties of the progenitor tumor. Retrotransposition-associated DNA damage and/or replication stress could be a potential molecular vulnerability for precision cancer medicine approaches.</p
Effects of a structured exercise program on motivational outcomes in patients with Colon cancer.
No full textBACKGROUND: CHALLENGE was the first phase 3 trial to examine the effects of exercise on cancer-related survival and, therefore, required a substantial and sustained increase in moderate-to-vigorous physical activity (MVPA) that had not been achieved in previous trials. Here, we report the effects of the intervention on the social cognitive beliefs that were targeted to achieve long-term exercise behavior change. METHODS: Patients with resected colon cancer who had completed chemotherapy were randomized to health education materials (HEM) or a structured exercise program (SEP) consisting of 48 behavioral support sessions delivered over a 3-year period. Social cognitive constructs from the Theory of Planned Behavior were assessed by single items using 5-point scales at baseline and every 6 months during the 3-year intervention. Regression models for repeated measurements were used to estimate the least square means and robust standard errors for each randomized group at each time point. RESULTS: Between 2009 and 2024, 889 patients were randomized to SEP (n = 445) or HEM (n = 444). SEP compared to HEM reported significantly more favorable social cognitive beliefs about exercise at almost all time points. Average intervention effects (AIE) across the 3-year intervention favored SEP for perceived benefit (AIE = 0.29; p </p
Precise Extraction of Croplands from Remote Sensing Images in Egypt by a Dual-Encoder U-Net with Multi-Scale Axial Attention and Boundary Constraints
No full textAccurate cropland parcel mapping is essential for food security and sustainable land management in arid Africa, yet it remains challenging in Egypt due to edge blurring, spectral confusion, and fragmented fields in medium-resolution imagery. A novel dual-encoder deep learning method that integrates multi-scale axial attention and boundary constraints (MAA-BCNet) is proposed for the precise extraction of croplands in Egypt from Sentinel-2 multispectral images. A dual-path encoder is designed to fuse CNN-based local textures with an RMT global branch using spatial decay attention for complementary feature extraction. A multi-scale axial attention module is introduced to capture anisotropic parcel structures for improved spectral–spatial discrimination, and a multi-directional gradient edge enhancement module is developed for explicitly preserving boundary integrity. A U-Net++ decoder is employed for dense multi-scale aggregation. Experimental results in Egypt demonstrate that MAA-BCNet achieves superior performance in delineating cropland parcels, particularly for irregular or fragmented croplands with complex landscapes and fuzzy boundaries. Compared with the widely used segmentation models such as DeepLabV3_plus, PSPnet, Link_net, FCN_resnet101, and U-Net++ under the same training and evaluation settings, our model has the best performance, with Recall, Precision, IoU, and F1-Score reaching 94.92%, 90.77%, 86.57%, and 92.80%, respectively. These advancements make MAA-BCNet suitable for cropland mapping of large areas of Egypt, with applications in precision agriculture and sustainable land management.</p
Part-Aware Cross-Integration Transformer with Proximity Guided Regularization for Domain Generalizable Animal Re-Identification
No full textAnimal re-identification (ReID) plays an essential role in large-scale wildlife monitoring, ecological research, and conservation management. However, unlike human ReID, animal ReID faces severe challenges stemming from inter-species diversity, irregular poses, and the lack of consistent anatomical landmarks. These difficulties are further amplified under the domain generalization setting–Re-identification of Any Animal in the Wild (ReID-AW), where the model must correctly identify unseen species. We propose the Part-Aware Cross-Integration Transformer (PACIT), a unified framework designed for multi-species animal ReID. PACIT contains two complementary modules. The Global-Part Cross Attention (GP-CA) module establishes structured interactions between global semantics and learnable part-aware tokens, guiding the model to attend to discriminative body regions under varying species morphologies. In parallel, the Proximity-Guided Regularization (PGR) module employs a memory-driven proxy task to mine cross-instance similarity beyond mini-batch constraints, enabling the joint exploitation of fine-grained identity cues and coarse-grained inter-species relationships. Extensive evaluations on ReID-AW benchmark are conducted under a unified protocol. PACIT achieves an average mAP of 43.3% and CMC-1 of 63.0%, consistently outperforming state-of-the-art approaches. We believe PACIT provides a scalable foundation for more universal cross-species identification systems in real-world ecological applications.</p