Search or filter publications

Filter by type:

Filter by publication type

Filter by year:

to

Results

  • Showing results for:
  • Reset all filters

Search results

  • Journal article
    Zhao Y, Chen Q, Li H, Zhou H, Attar HR, Pfaff T, Wu T, Li Net al., 2026,

    Recurrent U-Net-based Graph Neural Network (RUGNN) for accurate deformation predictions in sheet material forming

    , Advanced Engineering Informatics, Vol: 69, ISSN: 1474-0346

    In recent years, various artificial intelligence-based surrogate models have been proposed to provide rapid manufacturability predictions of material forming processes. However, traditional AI-based surrogate models, typically built with scalar or image-based neural networks, are limited in their ability to capture complex 3D spatial relationships and to operate in a permutation-invariant manner. To overcome these issues, emerging graph-based surrogate models are developed using graph neural networks. This study developed a new graph neural network surrogate model named Recurrent U Net-based Graph Neural Network (RUGNN). The RUGNN model can achieve accurate predictions of sheet material deformation fields across multiple forming timesteps. The RUGNN model incorporates Gated Recurrent Units (GRUs) to model temporal dynamics and a U-Net inspired graph-based downsample/upsample mechanism to handle spatial long-range dependencies. A novel ’node-to-surface’ contact representation method was proposed, offering significant improvements in computational efficiency for large-scale contact interactions. The RUGNN model was validated using a cold forming case study and a more complex hot forming case study using aluminium alloys. Results demonstrate that the RUGNN model provides accurate deformation predictions closely matching ground truth FE simulations and outperforming several baseline GNN architectures. Model tuning was also performed to identify suitable hyperparameters, training strategies, and input feature representations. These results demonstrate that RUGNN is a reliable approach to support sheet material forming design by enabling accurate manufacturability predictions.
  • Journal article
    Tu Y, Wu B, Martínez-Pañeda E, 2026,

    Phase field modelling of cracking and capacity fade in core-shell cathode particles for lithium-ion batteries

    , Applied Energy, Vol: 403, Pages: 127105-127105, ISSN: 0306-2619
  • Journal article
    Baxter W, Echavarri IV, Porat T, 2025,

    Exploring the impact of changing government policy on vaccination eligibility for 50-64 year olds: A qualitative thematic analysis in England and Scotland

    , HUMAN VACCINES & IMMUNOTHERAPEUTICS, Vol: 21, ISSN: 2164-5515
  • Journal article
    Chan EYK, Yu X, Qin C, Ghajari Met al., 2025,

    Balancing efficiency and accuracy: extreme gradient boosting and neural networks for near real-time brain deformation prediction in sports collisions

    , Engineering Applications of Artificial Intelligence, Vol: 162, ISSN: 0952-1976

    Rapid head motion during sports collisions can cause traumatic brain injury. Head motion can be measured with instrumented mouthguards and fed into finite element (FE) models to predict brain strain, a measure of brain deformation and injury. Due to the computational cost of FE models, deep neural networks have been developed for near real-time prediction. However, they are not used in pitch-side assessments due to their complexity and reliance on full kinematic data, which cannot be reliably transmitted in real-time.We propose an extreme gradient boosting (XGBoost) model with simple input of two kinematic features. Its accuracy and efficiency were compared with two deep learning models: a multilayer perceptron (MLP) using 20 features, and a convolutional neural network (CNN) using entire kinematics. All models were trained on 1701 rugby impacts collected with mouthguards and simulated using the Imperial brain FE model. The XGBoost model predicted strain in key brain regions, while the deep learning models predicted whole-brain strain distributions.All models showed reasonable accuracy in predicting regional strain, with R2 values 0.764–0.851 for XGBoost, 0.721–0.876 for MLP, and 0.744–0.887 for CNN. XGBoost required orders of magnitude fewer floating-point operations, and it used simple input that can be calculated on mouthguards and reliably transmitted in real-time.This study suggests that different models can be used at different stages of brain injury assessment. We hope that the XGBoost model proposed here will lower the barriers for adopting brain strain combined with instrumented mouthguards for pitch-side assessments from elite to grassroot collision sports.
  • Journal article
    Hana Frade JL, Engracia Giraldi JDM, Porat T, 2025,

    The influence of national origin cues in HPV vaccination advertising: an eye-tracking study of visual attention and vaccine perception using quantitative and qualitative analysis

    , Human Vaccines & Immunotherapeutics, Vol: 21, ISSN: 2164-5515

    This study is among the first to investigate how national origin cues influence visual attention and perception in HPV vaccine advertisements, using eye-tracking technology to provide objective insights into consumer responses. By integrating methods from public health, psychology, and advertising research, this study explores how visual attention is shaped by national affiliation cues. In a controlled experimental setting with a sample of 40 UK university students, we investigated visual attention and effectiveness of HPV vaccination advertisements by comparing ads disclosing the national origin of the vaccine and without any origin information. We assessed total fixation duration and time to first fixation to various elements of the ad, along with intention and attitude measures. Contrary to one of our hypotheses, we did not find significant differences in intention (p = .758) and attitude (p = .620) measures. However, there was significant difference in total fixation duration toward one of the ad images between conditions (p = .043). The qualitative analysis reveals the role of country-of-origin (COO) in HPV vaccination advertising, suggesting a shift in attention from that image to the COO cue. Furthermore, eight out of the 20 participants in the treatment condition did not fixate at the COO cue. Findings provide critical insights for public health communication strategies, suggesting that the use (or omission) of national origin cues in vaccine advertisements could influence vaccine perception and hesitancy. These results highlight the need for strategic messaging approaches to enhance HPV vaccine acceptance and improve public trust in domestic and international vaccines.
  • Journal article
    Kharman AM, Jursitzky C, Zhou Q, Ferraro P, Marecek J, Pinson P, Shorten Ret al., 2025,

    An adversarially robust data market for spatial, crowd-sourced data

    , Distributed Ledger Technologies: Research and Practice, Vol: 4, Pages: 1-20, ISSN: 2769-6472

    We describe an architecture for a decentralised data market for applications in which agents are incentivised to collaborate to crowd-source their data. The architecture is designed to reward data that furthers the market’s collective goal, and distributes reward fairly to all those that contribute with their data. We show that the architecture is resilient to Sybil, wormhole and data poisoning attacks. In order to evaluate the resilience of the architecture, we characterise its breakdown points for various adversarial threat models in an automotive use case.
  • Journal article
    Batcup C, Almukhtar A, Menon A, Leff D, Judah G, Demirel P, Porat Tet al., 2025,

    Barriers and enablers to sustainable anaesthetic practice: a mixed-methods study

    , British Journal of Anaesthesia, ISSN: 0007-0912

    Background: Anaesthetic practices contribute significantly to the environmental impact of healthcare. Using local or regional anaesthesia instead of general anaesthesia, and TIVA instead ofinhalational anaesthesia, can reduce this impact. This study investigated why general anaesthesia is sometimes used over local and regional anaesthesia, and why inhalational agents are often chosen over TIVA.Methods: We conducted a mixed-methods study in the UK (June 2023–April 2024), underpinned by the Theoretical Domains Framework. Semi-structured interviews (n=19) with anaesthetists, surgeons and nurses of differing seniority were analysed using Framework Analysis. A national survey (n=347), distributed via posters and professional networks, was developed from early interview findings. Quantitative data were analysed descriptively and open-text responses were coded using the qualitative framework.Results: Four key themes were identified: (1) contextual factors affecting anaesthesia decision making; (2) patient differences and preferences; (3) influence of key decision makers on anaesthesia choice; and (4) default practices and lack of confidence in alternatives. These encompassed 17subthemes and mapped to 9 of 14 Theoretical Domains Framework domains.Conclusions: This study provides new insights into behavioural influences underlying anaesthetic practice, which can inform the design of interventions to improve the sustainability of anaesthesia, without compromising patient safety and comfort. Addressing systemic and behavioural barriers through dedicated local anaesthesia operating lists, improved patient communication, targeted training and supportive technologies may enhance efficiency while promoting safe, sustainable, patient-centred practice. Future interventions should be co-designed with surgeons, anaesthetists and patients to ensure clinical acceptability, feasibility, and sustainability
  • Journal article
    Almukhtar A, Batcup C, Jagannath S, Leff D, Porat T, Judah G, Demirel Pet al., 2025,

    Understanding sustainability in operating theatres: an ethnographic study to determine drivers of unsustainable behaviours

    , Annals of Surgery Open, ISSN: 2691-3593

    BackgroundClimate change is the biggest threat to human health. Paradoxically, the healthcare sector is a major contributor to climate change, and operating theatres are among the highest sources of emissions. Unsustainable practices are actions that compromise environmental, social, and financial sustainability, leading to unnecessary resource use, avoidable harm to the wider population, and reduced ability to provide effective healthcare in the future. Drivers of unsustainable practices and barriers to sustainability in practice (a top priority identified by the James Lind Alliance Priority Setting Partnership) are unexplored, hindering interventions which can help meet net-zero targets within healthcare. We conducted the first known ethnographic study to investigate behaviours related to sustainability in operating theatres, and their influences on those behaviours to inform the design of effective behaviour change interventions.MethodsNon-participant ethnographic observations with opportunistic discussions in elective general surgical operating theatres were conducted between June and December 2023 at two university hospitals in Central London. Data were collected until saturation using a template developed during the initial observations. Inductive thematic analysis was conducted, with sub-themes (influences) deductively mapped to the Theoretical Domains Framework (TDF).ResultsTwenty-six procedures were observed (42 hours). Unsustainable behaviours included: (i) unnecessary and inappropriate glove use, potentially compromising safety (average 8-10 pairs per operation), (ii) incorrect waste disposal, (iii) unnecessary package opening, and (iv) energy waste. Thematic analysis generated 6 themes and 16 influences (mapped to 9 TDF domains). Key themes were that sustainable practices are “infrequent and inconsistent” due to limited awareness (Knowledge) and low environmental concerns (Memory, Attention and Decision Processes). Unsustainable behaviours we
  • Journal article
    Liu H, Luo W, Xu M, Zhang S, Xue J, Chen Q, Zhao Y, Hu N, Gao Zet al., 2025,

    Emerging bioelectronics and optogenetics for neurogastroenterology

    , BIOSENSORS & BIOELECTRONICS, Vol: 288, ISSN: 0956-5663
  • Journal article
    Hamdan S, Pinson P, Treluyer J-M, Kaguelidou F, Chouchana L, Smail-Faugeron Vet al., 2025,

    Trends of Opioids and Nefopam Use Among French Children: A Population-Based Study From 2009 to 2023

    , CLINICAL DRUG INVESTIGATION, ISSN: 1173-2563
  • Journal article
    Koczias C, Demirel P, 2025,

    The impact of green skills and green capabilities on firms’ financial performance: a systematic literature review

    , Journal of Innovation & Knowledge, Vol: 10, ISSN: 2444-569X

    Green skills and capabilities in organisations are important enablers of net-zero transition as they determine firms’ ability to adopt sustainability as a core business strategy. This study conducts a systematic literature review of 84 scholarly articles and 7 industry publications to examine the conceptualisation of green skills and green capabilities in firms, their influence on firms’ financial performance, and the factors shaping these complex relationships. Findings reveal that the green skills and green capabilities concepts are broad and ambiguously defined, necessitating greater conceptual clarity, especially with respect to how the green skills and capabilities relate to each other. Accordingly, the study develops a framework outlining different green skill and capability typologies and their interconnections. Additionally, six key performance categories—(1) costs, (2) profitability, (3) efficiency and productivity, (4) firm growth, (5) liquidity, and (6) market performance— are identified inductively, with evidence from the literature indicating that both green skills and green capabilities positively impact each of these areas. Results suggest that while the overall effect of green skills and capabilities on financial performance is positive, several contextual factors, such as firm size, sector, and region, influence this relationship. The study highlights the most significant caveats and their implications, contributing to a more nuanced understanding of the knowledge–performance link for the net-zero transition.
  • Journal article
    Boufelja Y S, Quinn A, Shorten R, 2025,

    Randomized transport plans via hierarchical fully probabilistic design

    , Information Sciences, Vol: 718, ISSN: 0020-0255

    An optimal randomized strategy for design of balanced, normalized mass transport plans is developed. It replaces—but specializes to—the deterministic, regularized optimal transport (OT) strategy, which yields only a certainty-equivalent plan. The incompletely specified—and therefore uncertain—transport plan is acknowledged to be a random process. Therefore, hierarchical fully probabilistic design (HFPD) is adopted, yielding an optimal hyperprior supported on the set of possible transport plans, and consistent with prior mean constraints on the marginals of the uncertain plan. This Bayesian resetting of the design problem for transport plans—which we call HFPD-OT—confers new opportunities. These include (i) a strategy for the generation of a random sample of joint transport plans; (ii) randomized marginal contracts for individual source-target pairs; and (iii) consistent measures of uncertainty in the plan and its contracts. An application in fair market matching is outlined, in which HFPD-OT enables the recruitment of a more diverse subset of contracts—than is possible in classical OT—into the delivery of an expected plan.
  • Conference paper
    ZHOU J, Aloufi R, Porat T, Van Zalk Net al., 2025,

    Human-AI collaboration in high-stakes decision-making: work in progress

    , 38th International BCS Human-Computer Interaction Conference (BCS HCI 25), Publisher: BCS Learning and Development Ltd, Pages: 465-470

    This work in progress investigates human interaction with an LLM-powered chatbot, presented as either a fellow human or a transparently disclosed AI collaborator, in a high-stakes decision-making simulation—the NASA Moon Survival Task. We will employ a one-way between-subjects design to examine how individuals’ collaboration and communication are influenced by the identity of their partner (AI vs. human). Specifically, we will evaluate individuals’ collaboration processes (i.e., collaborative behaviour and communicative dynamics) and outcomes, alongside their retrospective interaction experience andperceptions of the partner. We will also examine dyadic-level linguistic coordination during the interaction and conduct user profiling to uncover variations in AI collaborative benefits. We anticipate that this studywill have four key impacts: safeguarding human-AI collaboration, democratising AI benefits, guiding modelimprovement, and making methodological contributions. The anonymised dialogues and associated data will be open-sourced upon study completion.
  • Journal article
    Cardenas MK, Anza-Ramirez C, Bernabe-Ortiz A, Custodio N, Montesinos R, Miranda JJ, Da Re M, Belon-Hercilla MV, Lazo-Porras M, Hawkins J, Diez-Canseco F, Moore G, Whiteley W, Calvo RA, Cuba-Fuentes MS, Landeiro F, Butler CRet al., 2025,

    Validation and cost-effectiveness of an mHealth tool for cognitive impairment detection in Peru: protocol for the IMPACT Salud observational study

    , BMJ Open, Vol: 15, ISSN: 2044-6055

    Introduction Dementia is a chronic and progressive neurological condition characterised by cognitive and functional impairment. It is often associated with multimorbidity and imposes a significant economic burden on healthcare systems and families, especially in low-income and middle-income countries. In Peru, where dementia cases are increasing rapidly, timely detection and referral for diagnosis is crucial. This protocol is part of the IMPACT Salud project in Peru. Here, we focus on a specific component aimed at validating an mHealth tool for the detection of cognitive and functional impairment and assessing its cost-effectiveness. We will also assess changes in cognitive and functional impairment as well as health economic outcomes over 1 year.Methods and analysis This observational study will be conducted in four geographically diverse regions of Peru. Community health workers are expected to contact approximately 32 000 participants (≥60 years) to apply an mHealth-enabled tool that includes cognitive and functional instruments: Ascertain Dementia 8, Peruvian version of Rowland Universal Dementia Assessment Scale and Pfeffer Functional Activities Questionnaire. From this large sample, we aim to find 3600 participants and their study partners to enrol and interview at baseline regarding sociodemographic characteristics, lifestyles, comorbidities and health economic data including resource use, costs and health-related quality of life (HR-QoL). Psychologists, blind to previous results, will assess dementia stage of the participants using an abbreviated Clinical Dementia Rating (CDR) scale. At 6-month follow-up, participants will complete a brief health economics questionnaire on resource use, costs and HR-QoL. To validate the accuracy of the detection tool, a subsample of 600 participants who completed the baseline will undergo a gold-standard clinical neuropsychological assessment. This subsample will participate in a 12-month follow-up, including healt
  • Journal article
    Qi N, Pinson P, Almassalkhi MR, Zhuang Y, Su Y, Liu Fet al., 2025,

    Capacity credit evaluation of generalized energy storage considering strategic capacity withholding and decision-dependent uncertainty

    , APPLIED ENERGY, Vol: 397, ISSN: 0306-2619
  • Journal article
    Jin B, Kobayashi A, Bhattacharya D, Seino A, Tokuda F, Tien NC, Kosuge Ket al., 2025,

    Automated Straight-Line Sewing of Stretchable Fabrics With Different Lengths

    , IEEE ROBOTICS AND AUTOMATION LETTERS, Vol: 10, Pages: 12165-12172, ISSN: 2377-3766
  • Journal article
    Ikeya K, Cardin M-A, Cilliers J, Starr SO, Hadler Ket al., 2025,

    Multi-objective decision-making for flexible design and planning of oxygen production facilities in an uncertain lunar environment

    , ACTA ASTRONAUTICA, Vol: 236, ISSN: 0094-5765
  • Journal article
    Lai H, Soreq E, Bourke N, Baker C, Zimmerman K, Parkinson M, Daniels S, Gregg E, Sharp D, Li Let al., 2025,

    Traumatic brain injury (TBI) and mortality in older adults with and without pre-injury dementia

    , Age and Ageing, ISSN: 0002-0729

    Background: Incidence of traumatic brain injury (TBI) is rising in older adults. Dementia is a common comorbidity and may worsen post-TBI outcomes, but its effects have not been studied. Objective: To compare all-cause mortality following TBI or non-TBI trauma (NTT) and quantify the impacts of age, deprivation, and dementia.Design: Population-based retrospective cohort studySetting: Linked primary and secondary care electronic health records (EHRs) from the Secure Anonymised Information Linkage (SAIL) DatabankSubjects: Adult residents in Wales aged 18-100 with hospitalised TBI or NTT recorded between January 2000-December 2022. Methods: Cox proportional hazard models were used to estimate survival within 1, 6, and 12 months of hospitalised TBI/NTT in those with and without pre-injury dementia diagnosis. Groups were propensity-matched by age, sex, and morbidities. Models were stratified by age, sex, and deprivation.Results: 23,428 TBIs (n=18,940) and 589,169 NTTs (n=421,259) were identified. TBIs were associated with higher mortality than NTT at all timepoints. Older age was associated with higher mortality after TBI, with 16.9% one-month mortality in patients aged 65-79, and 31% in patients aged 80-100.In TBI patients, 30-day mortality was significant irrespective of dementia. 6- and 12-month mortality were higher in those with than without pre-injury dementia diagnosis. Conclusions: TBI is associated with higher all-cause mortality than NTT, particularly in older age. Among those with TBI, patients with pre-injury dementia had particularly high chronic mortality. There is an urgent need to understand the reasons for poor outcomes in older adult TBI populations, especially those with dementia.
  • Journal article
    Lu X, Owen RE, Du W, Zhang Z, Bertei A, Soni R, Zhang X, Iacoviello F, Li D, Llewellyn A, Chen J, Zhang H, Yao X, Li Q, Zhao Y, Marathe S, Rau C, Shearing PRet al., 2025,

    Unravelling electro-chemo-mechanical processes in graphite/silicon composites for designing nanoporous and microstructured battery electrodes

    , NATURE NANOTECHNOLOGY, ISSN: 1748-3387
  • Journal article
    ZHOU J, Van Zalk N, 2025,

    A person-oriented approach to social anxiety and depression: latent profiles and emotional functioning in adults

    , Frontiers in Psychology, Vol: 16, ISSN: 1664-1078

    Objectives: Symptoms of social anxiety and depression often co-occur, but many questions remain about symptom-level co-occurrence and the heterogeneity of symptom presentations across individuals, as well as their emotional functioning. This study aimed to investigate the co-occurrence of social anxiety and depressive symptoms in adults and variations in emotional functioning linking symptom heterogeneity. Methods: This study used a person-oriented approach, Latent Profile Analysis (LPA), to identify distinct profiles (i.e., subgroups) in a UK adult sample (N = 222) varying in presentations of social anxiety and depressive symptoms. Further analyses examined between-profile differences in emotional functioning, including daily affect and emotion regulation. Results: Four profiles were identified: Comorbid (12.61%), Dysphoric (10.36%), Socially Anxious (36.94%), and Low Distress (40.09%), replicating the four-profile solution revealed in prior research on adolescents. The Comorbid subgroup reported the most pronounced emotional dysfunction, with higher daily negative affect, lower positive affect, and greater emotion dysregulation than the other three subgroups. The Low Distress subgroup reported the best emotional functioning. Conclusions: The cross-sectional study design restricts our ability to evaluate the long-term stability of the identified profiles. Nevertheless, this study illuminates the diverse ways social anxiety and depression intertwine, underscoring the necessity of transdiagnostic interventions that cater to a wide range of symptom patterns and emotional functioning.
  • Journal article
    Bevan PA, BanksLeite C, Kovac M, Lawson J, Picinali L, Sethi SSet al., 2025,

    Robotics‐assisted acoustic surveys could deliver reliable, landscape‐level biodiversity insights

    , Remote Sensing in Ecology and Conservation, ISSN: 2056-3485

    Terrestrial remote sensing approaches, such as acoustic monitoring, deliver finely resolved and reliable biodiversity data. However, the scalability of surveys is often limited by the effort, time and cost needed to deploy, maintain and retrieve sensors. Autonomous unmanned aerial vehicles (UAVs, or drones) are emerging as a promising tool for fully autonomous data collection, but there is considerable scope for their further use in ecology. In this study, we explored whether a novel approach to UAV-based acoustic monitoring could detect biodiversity patterns across a varied tropical landscape in Costa Rica. We simulated surveys of UAVs employing intermittent locomotion-based sampling strategies on an existing dataset of 26,411 h of audio recorded from 341 static sites, with automated detections of 19 bird species (n = 1819) and spider monkey (n = 2977) vocalizations. We varied the number of UAVs deployed in a single survey (sampling intensity) and whether the UAVs move between sites randomly, in a pre-determined route to minimize travel time, or by adaptively responding to real-time detections (sampling strategy), and measured the impact on downstream ecological analyses. We found that avian species detections and spider monkey occupancy were not impacted by sampling strategy, but that sampling intensity had a strong influence on downstream metrics. Whilst our simulated UAV surveys were effective in capturing broad biodiversity trends, such as spider monkey occupancy and avian habitat associations, they were less suited for exhaustive species inventories, with rare species often missed at low sampling intensities. As autonomous UAV systems and acoustic AI analyses become more reliable and accessible, our study shows that combining these technologies could deliver valuable biodiversity data at scale.
  • Journal article
    Bhattacharya D, Cheung TK, Wang Y, Lau Det al., 2025,

    Kinematic and dynamic modeling of cable-object interference and wrapping in complex geometrical-shaped cable-driven parallel robots

    , Mechanism and Machine Theory, Vol: 214, ISSN: 0094-114X

    Cable-Driven Parallel Robots (CDPRs) use cables as actuators to maneuver rigid mobile-platform in a parallel mechanism setup. Typically, CDPR kinematic and dynamic models avoid cable-object (cable-mobile-platform and cable-obstacle) interferences to prevent sudden cable tension changes that could deviate the end-effector’s trajectory. However, allowing these interferences can lead to cable wrapping, where cables wrap around complex-shaped surfaces upon contact, enhancing the CDPR’s workspace and reducing its footprint. Despite the potential benefits, there currently exists no kinematic and dynamic model that effectively incorporates cable wrapping around such complex-shaped surfaces. This paper introduces a novel numerical-based kinematic and dynamic modeling framework for CDPRs that detects and then manages cable wrapping around mobile-platform and multiple obstacles with the assumption that the cables remain taut and for every position along the cable, there is a unique and smooth way to describe its location on the surface. Simulation and hardware results on various complex-shaped mobile-platform and obstacles show that the proposed model framework can be conveniently and effectively applied to the real-time modeling of cable wrapping. Code and videos available at: https://github.com/bhattner143/GeoWrapSim-CDPR.git.
  • Journal article
    Lei G, Cooper SJ, 2025,

    Do Llamas understand the periodic table?

    , Digital Discovery, ISSN: 2635-098X

    Large Language Models (LLMs) demonstrate remarkable abilities in synthesizing scientific knowledge, yet their limitations, particularly with basic arithmetic, raise questions about their reliability. As materials science increasingly employs LLMs for tasks like hypothesis generation, understanding how these models encode specialized knowledge becomes crucial. Here, we investigate how the open-source Llama series of LLMs represent the periodic table of elements. We observe a 3D spiral structure in the hidden states of LLMs that aligns with the conceptual structure of the periodic table, suggesting that LLMs can reflect the geometric organization of scientific concepts learned from text. Linear probing reveals that middle layers encode continuous, overlapping attributes that enable indirect recall, while deeper layers sharpen categorical distinctions and incorporate linguistic context. These findings suggest that LLMs represent symbolic knowledge not as isolated facts, but as structured geometric manifolds that intertwine semantic information across layers. We hope this inspires further exploration into the interpretability mechanisms of LLMs within chemistry and materials science, enhancing trust of model reliability, guiding model optimization and tool design, and promoting mutual innovation between science and AI.
  • Journal article
    Kormushev P, 2025,

    Rescue robots for casualty extraction: a comprehensive review

    , IEEE Access, ISSN: 2169-3536

    The development of robotic systems for search and rescue (SAR) operations holds greatpotential in reducing risks to rescue workers and enhancing the likelihood of successful rescue operations. Although many studies have been conducted on search and rescue robotics, more emphasis should be placed on developing rescue robots, especially those intended for physical rescue interventions, such as loading and transporting injured individuals (i.e., casualties) to a secure area—the process known as casualty extraction. The enabling technologies for such robots remain challenging due to the complexity of the tasks, significantly high safety considerations, and strict standards—since the robots are designed to interact with casualties physically. In addition to academic institutions, most research and technical implementations of state-of-the-art casualty extraction robots are carried out by military organisations. This paper presents a comprehensive review of the current state of the art in developing mobile rescue robots for casualty extraction. The existing casualty extraction robot proposals publicly available in the literature are discussed and evaluated in terms of their design and morphology. Moreover, this review details the proposed casualty extraction procedure that corresponds to the robot designs, the operation method, and the levels of autonomy of these robots. The existing state-of-the-art technologies are discussed and compared to evaluate the pros and cons of each system, providing a guideline for further research into areas where more effort could be applied. Based on the review and evaluation of the existing state-of-the-art, we identify critical research gaps that require further investigation to enhance the current state of the art and facilitate long-term development in rescue robotics research.
  • Journal article
    Missoni F, Poole KC, Picinali L, Canessa Aet al., 2025,

    Effects of auditory distance cues and reverberation on spatial perception and listening strategies

    , npj Acoustics, Vol: 1, ISSN: 3005-141X

    Spatial hearing—the brain’s ability to identify sound origins using auditory cues—is inherently multisensory, integrating vision, hearing, and proprioception to reduce uncertainty and support adaptive interaction with the environment. While simplified experimental paradigms have advanced our understanding, their limited ecological validity limits real-world applicability. This study investigates how listener movement, reverberation, and distance affect localisation accuracy in more naturalistic settings. Participants performed an active localisation task without prescribed listening strategies, in either anechoic or reverberant conditions. Sound sources were positioned around them in both horizontal and vertical planes, at varying distances. Results show increased head movement in reverberant environments, suggesting an adaptive response to degraded binaural cues. While distance did not influence listening strategies, it significantly affected localisation accuracy. These findings highlight the importance of considering ecological factors when studying spatial hearing and suggest that natural listener behaviour plays a key role in maintaining spatial accuracy under different reverberant conditions.
  • Journal article
    Rakotomalala Robinson D, Pennisi I, Cavuto M, Kiemde F, Chamai M, Yrgnur Some D, Quigley E, Malpartida Cardenas K, Ousmane Ndiath M, Correa S, Darboe B, Stewart L, Georgiou P, Baldeh M, Tinto H, Cunnington A, Erhart A, DAlessandro U, Rodriguez Manzano J, On behalf of the NIHR Global Health Research Group on Digital Diagnostics for African Health Systemset al., 2025,

    Sensitive near point-of-care detection ofasymptomatic and sub-microscopicPlasmodium falciparum infections in Africanendemic countries

    , Nature Communications, ISSN: 2041-1723

    Limited diagnostic capacity for detecting asymptomatic malaria infections with low parasite densities hinders elimination efforts in Africa. Here, we adapt a near point-of-care, LAMP-based diagnostic platform for malaria diagnosis using capillary blood. This Pan/Pf detection method meets the Malaria Eradication Research Agenda (malERA) criteria for community-level screening, witha limit of detection of 0.6 parasites/μL and a sample-to-result time under 45 minutes. We evaluate its performance on 672 capillary blood samples collected at the community level in The Gambia and Burkina Faso, including 146Plasmodium falciparum positives confirmed by qPCR. The diagnostic platform achieves 95.2% sensitivity (95% CI: 90.4–98.1) and 96.8% specificity (95% CI: 94.9–98.0). It also detects 94.9% (130/137) of asymptomatic infections and 95.3% (41/43) of sub-microscopic cases (<16 parasites/μL), outperforming expert microscopy (70.1% and 0%) and rapid diagnostic tests (49.6% and 4.7%). This field-deployable molecular diagnostic method offers a sensitive, scalable solution to support test-and-treat strategies for malaria elimination across Africa.
  • Journal article
    Kalossaka LM, Mohammed AA, Bastos L, Barter LMC, Myant CWet al., 2025,

    Green light vat‐photopolymerisation for 3D printing hydrogels with complex lattice structures

    , Advanced Materials Technologies, Vol: 10, ISSN: 2365-709X

    Moving beyond UV curing systems opens new potential application spacessuch as biological, portable printing solutions, as well as innovative chemistriesand material properties. A novel visible light printer is proposed for the first timeusing green Digital Light Processing (gDLP) at a wavelength of 514 nm. GreenLED lights are integrated into a commercial desktop DLP printer to 3D printhydrogels with complex designs at high resolution. A workflow process is pre-sented to develop and optimize formulations for gDLP, resulting in two novel in-house photoresin formulations made specifically for green light printing. Theseformulations comprise PEGDA 700 with and without acrylamide, using a type IIphotoinitiating system of Eosin Y, triethylamine, and N-vinylpyrrolidone. Thephotoresins are optimized to achieve highly vascularized lattice prints by mod-ulating layer light exposure, chemical components, and photoinitiator concen-trations. The gDLP successfully printed hydrogels with a layer height of 50 𝛍mand feature dimensions as small as 0.3 mm by adjusting light duration per layer.3D printed hydrogels using both formulations are tested for varying designcomplexity, including ISO/ASTM standards, and evaluated with optical imag-ing, SEM, and mechanical testing. This study highlights gDLP technology’s po-tential for diverse applications in tissue engineering and sustainable materials.
  • Journal article
    Yin Y, Zuo H, Jennings T, Sandeep J, Cartwright B, Buhagiar J, Williams P, Adams K, Hazeri K, Childs Pet al., 2025,

    Use and potential of AI in assisting surveyors in building retrofit and demolition - a scoping review

    , Buildings, Vol: 15, ISSN: 2075-5309

    Background: Pre-retrofit auditing and pre-demolition auditing (PRA/PDA) are important in material reuse, waste reduction, and regulatory compliance in the building sector. An emphasis on sustainable construction practices has led to a higher requirement for PRA/PDA. However, traditional auditing processes demand substantial time and manual effort and are more easily to create human errors. As a developing technology, artificial intelligence (AI) can potentially assist PRA/PDA processes. Objectives: This scoping review aims to review the potential of AI in assisting each sub-stage of PRA/PDA processes. Eligibility Criteria and Sources of Evidence: Included sources were English-language articles, books, and conference papers published before 31 March 2025, available electronically, and focused on AI applications in PRA/PDA or related sub-processes involving structured elements of buildings. Databases searched included ScienceDirect, IEEE Xplorer, Google Scholar, Scopus, Elsevier, and Springer. Results: The review indicates that although AI has the potential to be applied across multiple PRA/PDA sub-stages, actual application is still limited. AI integration has been most prevalent in floor plan recognition and material detection, where deep learning and computer vision models achieved notable accuracies. However, other sub-stages—such as operation and maintenance document analysis, object detection, volume estimation, and automated report generation—remain underexplored, with no PRA/PDA specific AI models identified. These gaps highlight the uneven distribution of AI adoption, with performance varying greatly depending on data quality, available domain-specific datasets, and the complexity of integration into existing workflows. Conclusions: Out of multiple PRA/PDA sub-stages, AI integration was focused on floor plan recognition and material detection, with deep learning and computer vision models achieving over 90% accuracy. Other stages such as operation
  • Journal article
    Marggraf-Turley N, Pontoppidan NH, Picinali L, 2025,

    The impact of individual head-related transfer function augmentation on spatial release from masking.

    , Hear Res, Vol: 466

    Binaural hearing underpins effective communication in complex acoustic environments by increasing listeners' abilities to segregate concurrent sound sources. In certain conditions, interaural magnification of binaural cues has been shown to improve speech intelligibility in competing target masker scenarios, yet existing methods primarily comprise hearing aid algorithms, which, due to processing constraints, cause unwanted artefacts. Moreover, the perceptual effects of applying interaural magnification directly to a person's own head-related transfer function (HRTF) remain unclear. We revisit a previous method for magnifying interaural cues, apply it to individual HRTFs, and propose a modified version designed to reduce spatial distortions. We evaluate both approaches using auditory models and listening tests involving speech-on-speech masking with a midline target and symmetrical maskers. Behavioural results show that although magnifying interaural phase and level differences improves overall speech intelligibility compared to unmodified HRTFs, it does not significantly enhance spatial release from masking (SRM). Meanwhile, auditory model predictions suggest greater SRM gains than those observed behaviourally, suggesting an influence of listener familiarity with their individual spatial cues. Finally, interaural magnification benefits diminish for listeners who already exhibit strong SRM with their own HRTFs.
  • Journal article
    Deterding S, Guckelsberger C, Lintunen EM, Ady NMet al., 2025,

    Advancing Self-Determination Theory via computational modelling: the case of competence and optimal challenge

    , MOTIVATION AND EMOTION, ISSN: 0146-7239

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: http://www.imperial.ac.uk:80/respub/WEB-INF/jsp/search-t4-html.jsp Request URI: /respub/WEB-INF/jsp/search-t4-html.jsp Query String: id=1221&limit=30&resgrpMemberPubs=true&respub-action=search.html Current Millis: 1764903813547 Current Time: Fri Dec 05 03:03:33 GMT 2025

Contact us

Dyson School of Design Engineering
Imperial College London
25 Exhibition Road
South Kensington
London
SW7 2DB

design.engineering@imperial.ac.uk
Tel: +44 (0) 20 7594 8888

Campus Map