Search

Neural text-to-speech synthesis (TTS) captures prosodicfeatures strikingly well, notwithstanding the lack of prosodiclabels in training or synthesis. We trained a voice on a singleSwedish speaker reading in Swedish and English. The resultingTTS allows us to control the degree of English-accentedness inSwedish sentences. English-accented Swedish commonlyexhibits well-known prosodic characteristics such as erroneoustonal accents and understated or missed durational differences.TTS quality was verified in three ways. Automatic speechrecognition resulted in low errors, verifying intelligibility.Automatic language classification had Swedish as the majoritychoice, while the likelihood of English increased with ourtargeted degree of English-accentedness. Finally, a rank ofperceived English-accentedness acquired through pairwisecomparisons by 20 human listeners demonstrated a strongcorrelation with the targeted English-accentedness.We report on phonetic and prosodic analyses of theaccented TTS. In addition to the anticipated segmentaldifferences, the analyses revealed temporal and prominencerelated variations coherent with Swedish spoken by Englishspeakers, such as missing Swedish stress patterns and overlyreduced unstressed syllables. With this work, we aim to gleaninsights into speech prosody from the latent prosodic featuresof neural TTS models. In addition, it will help implementspeech phenomena such as code switching in TTS, QC 20240705, Deep learning based speech synthesis for reading aloud of lengthy and information rich texts in Swedish (2018-02427), Språkbanken Tal (2017-00626)

Neural text-to-speech synthesis (TTS) captures prosodicfeatures strikingly well, notwithstanding the lack of prosodiclabels in training or synthesis. We trained a voice on a singleSwedish speaker reading in Swedish and English. The resultingTTS allows us to control the degree of English-accentedness inSwedish sentences. English-accented Swedish commonlyexhibits well-known prosodic characteristics such as erroneoustonal accents and understated or missed durational differences.TTS quality was verified in three ways. Automatic speechrecognition resulted in low errors, verifying intelligibility.Automatic language classification had Swedish as the majoritychoice, while the likelihood of English increased with ourtargeted degree of English-accentedness. Finally, a rank ofperceived English-accentedness acquired through pairwisecomparisons by 20 human listeners demonstrated a strongcorrelation with the targeted English-accentedness.We report on phonetic and prosodic analyses of theaccented TTS. In addition to the anticipated segmentaldifferences, the analyses revealed temporal and prominencerelated variations coherent with Swedish spoken by Englishspeakers, such as missing Swedish stress patterns and overlyreduced unstressed syllables. With this work, we aim to gleaninsights into speech prosody from the latent prosodic featuresof neural TTS models. In addition, it will help implementspeech phenomena such as code switching in TTS, QC 20240705, Deep learning based speech synthesis for reading aloud of lengthy and information rich texts in Swedish (2018-02427), Språkbanken Tal (2017-00626)

Neural text-to-speech synthesis (TTS) captures prosodicfeatures strikingly well, notwithstanding the lack of prosodiclabels in training or synthesis. We trained a voice on a singleSwedish speaker reading in Swedish and English. The resultingTTS allows us to control the degree of English-accentedness inSwedish sentences. English-accented Swedish commonlyexhibits well-known prosodic characteristics such as erroneoustonal accents and understated or missed durational differences.TTS quality was verified in three ways. Automatic speechrecognition resulted in low errors, verifying intelligibility.Automatic language classification had Swedish as the majoritychoice, while the likelihood of English increased with ourtargeted degree of English-accentedness. Finally, a rank ofperceived English-accentedness acquired through pairwisecomparisons by 20 human listeners demonstrated a strongcorrelation with the targeted English-accentedness.We report on phonetic and prosodic analyses of theaccented TTS. In addition to the anticipated segmentaldifferences, the analyses revealed temporal and prominencerelated variations coherent with Swedish spoken by Englishspeakers, such as missing Swedish stress patterns and overlyreduced unstressed syllables. With this work, we aim to gleaninsights into speech prosody from the latent prosodic featuresof neural TTS models. In addition, it will help implementspeech phenomena such as code switching in TTS, QC 20240705, Deep learning based speech synthesis for reading aloud of lengthy and information rich texts in Swedish (2018-02427), Språkbanken Tal (2017-00626)

Neural text-to-speech synthesis (TTS) captures prosodicfeatures strikingly well, notwithstanding the lack of prosodiclabels in training or synthesis. We trained a voice on a singleSwedish speaker reading in Swedish and English. The resultingTTS allows us to control the degree of English-accentedness inSwedish sentences. English-accented Swedish commonlyexhibits well-known prosodic characteristics such as erroneoustonal accents and understated or missed durational differences.TTS quality was verified in three ways. Automatic speechrecognition resulted in low errors, verifying intelligibility.Automatic language classification had Swedish as the majoritychoice, while the likelihood of English increased with ourtargeted degree of English-accentedness. Finally, a rank ofperceived English-accentedness acquired through pairwisecomparisons by 20 human listeners demonstrated a strongcorrelation with the targeted English-accentedness.We report on phonetic and prosodic analyses of theaccented TTS. In addition to the anticipated segmentaldifferences, the analyses revealed temporal and prominencerelated variations coherent with Swedish spoken by Englishspeakers, such as missing Swedish stress patterns and overlyreduced unstressed syllables. With this work, we aim to gleaninsights into speech prosody from the latent prosodic featuresof neural TTS models. In addition, it will help implementspeech phenomena such as code switching in TTS, QC 20240705, Deep learning based speech synthesis for reading aloud of lengthy and information rich texts in Swedish (2018-02427), Språkbanken Tal (2017-00626)

Neural text-to-speech synthesis (TTS) captures prosodicfeatures strikingly well, notwithstanding the lack of prosodiclabels in training or synthesis. We trained a voice on a singleSwedish speaker reading in Swedish and English. The resultingTTS allows us to control the degree of English-accentedness inSwedish sentences. English-accented Swedish commonlyexhibits well-known prosodic characteristics such as erroneoustonal accents and understated or missed durational differences.TTS quality was verified in three ways. Automatic speechrecognition resulted in low errors, verifying intelligibility.Automatic language classification had Swedish as the majoritychoice, while the likelihood of English increased with ourtargeted degree of English-accentedness. Finally, a rank ofperceived English-accentedness acquired through pairwisecomparisons by 20 human listeners demonstrated a strongcorrelation with the targeted English-accentedness.We report on phonetic and prosodic analyses of theaccented TTS. In addition to the anticipated segmentaldifferences, the analyses revealed temporal and prominencerelated variations coherent with Swedish spoken by Englishspeakers, such as missing Swedish stress patterns and overlyreduced unstressed syllables. With this work, we aim to gleaninsights into speech prosody from the latent prosodic featuresof neural TTS models. In addition, it will help implementspeech phenomena such as code switching in TTS, QC 20240705, Deep learning based speech synthesis for reading aloud of lengthy and information rich texts in Swedish (2018-02427), Språkbanken Tal (2017-00626)

The increasing anthropogenic CO2 forcing of the climate system calls for a better understanding of how polar ice sheets may respond to accelerating global warming. The sensitivity of the Greenland ice sheet to polar amplification, changes in ocean heat transport, and deteriorating perennial sea ice conditions makes the Northeast Greenland margin a pertinent location with respect to understanding the impact of climate change on ice sheet instability and associated sea level rise. Throughout the Cenozoic, ocean heat fluxes toward and along Northeast Greenland have been controlled by water mass exchanges between the Arctic and Atlantic oceans. A key element here is the current flow through oceanic gateways, notably the Fram Strait and the Greenland–Scotland Ridge. To gain a long-term (million-year) perspective of ice sheet variability in this region, it is essential to understand the broader context of ice–ocean–tectonic interactions. Coupling between the ice sheet, the subsurface, the ocean, and sea ice are readily observable today in Northeast Greenland, but geological records to illuminate long-term trends and their interplay with other parts of the global climate system are lacking. Consequently, the NorthGreen workshop was organized by the Geological Survey of Denmark and Greenland in collaboration with Aarhus (Denmark) and Stockholm (Sweden) universities in November 2022 to develop mission-specific platform (MSP) proposals for drilling the Northeast Greenland margin under the umbrella of the MagellanPlus Workshop Series Programme of the European Consortium for Ocean Research Drilling (ECORD). Seventy-one participants representing a broad scientific community discussed key scientific questions and primary targets that could be addressed through scientific drilling in Northeast Greenland. Three pre-proposals were initiated during the workshop targeting Morris Jesup Rise, the Northeast Greenland continental shelf, and Denmark Strait.

Silt-rich meltwater plume deposits (MPDs) analyzed from marine sediment cores have elucidated relationships that are clearly connected, yet difficult to constrain, between subglacial hydrology, ice-marginal landforms, and grounding-zone retreat patterns for several glacial catchments. Few attempts have been made to infer details of subglacial hydrology, such as flow regime, geometry of drainage pathways, and mode(s) of sediment transport through time, from grain-scale characteristics of MPDs. Using sediment samples from MPD, till, and grounding-zone proximal diamicton collected offshore of six modern and relict glacial catchments in both hemispheres, we examine grain shape distributions and microtextures (collectively, grain micromorphology) of the silt fraction to explore whether grains are measurably altered from their subglacial sources via meltwater action. We find that 75 % of all imaged grains (n = 9400) can be described by 25 % of the full range of measured shape morphometrics, indicating grain shape homogenization through widespread and efficient abrasive processes in subglacial environments. Although silt grains from MPDs exhibit edge rounding more often than silt grains from tills, grain surface textures indicative of fluvial transport (e.g., v-shaped percussions) occur in only a modest number of grains. Furthermore, MPD grain surfaces retain several textures consistent with transport beneath glacial ice (e.g., straight or arcuate steps, (sub)linear fractures) in comparable abundances to till grains. Significant grain shape alteration in MPDs compared to their till sources is observed in sediments from glacial regions where (1) high-magnitude, potentially catastrophic meltwater drainage events are inferred from marine sediment records and (2) submarine landforms suggest supraglacial melt contributed to the subglacial hydrological budget. This implies that quantifiable grain shape alteration in MPDs could reflect a combination of high-energy flow of subglac

DFT+U provides a convenient, cost-effective correction for the self-interaction error (SIE) that arises when describing correlated electronic states using conventional approximate density functional theory (DFT). The success of a DFT+U(+J) calculation hinges on the accurate determination of its Hubbard U and Hund J parameters, and the linear response (LR) methodology has proven to be computationally effective and accurate for calculating these parameters. This study provides a high-throughput computational analysis of the U and J values for transition metal d-electron states in a representative set of over 1000 magnetic transition metal oxides (TMOs), providing a frame of reference for researchers who use DFT+U to study transition metal oxides. In order to perform this high-throughput study, an atomate workflow is developed for calculating U and J values automatically on massively parallel supercomputing architectures. To demonstrate an application of this workflow, the spin-canting magnetic structure and unit cell parameters of the multiferroic olivine LiNiPO4 are calculated using the computed Hubbard U and Hund J values for Ni-d and O-p states, and are compared with experiment. Both the Ni-dU and J corrections have a strong effect on the Ni-moment canting angle. Additionally, including a O-pU value results in a significantly improved agreement between the computed lattice parameters and experiment.

A recently acquired multidisciplinary dataset comprising acoustic surveys (high-resolution sub- bottom profiles, multi-beam bathymetry, and broad band mid-water echo sounder), geochemistry (gas chemical and isotopic composition, porewater chemistry), and sedimentology (core lithology and X-ray CT) in the area of the Landsort deep (450 m of depth), south of Stockholm Archipelago, revealed the existence of an extensive (20 km2) region of the seafloor where massive gas release is occurring in the form of multiple bubble streams. This new discovery represents a major seafloor methane release site in Europe and is comparable in area to other large sites worldwide such as the ones in Svalbard and in the South Atlantic Ocean associated with gas hydrate provinces. The gas is formed mostly by methane of microbial origin. Surprisingly, bubbles rise 100’s of meters above the seafloor and reach surface waters above the halocline/oxycline at around 80 m of depth. Some bubbles appear to reach the sea-air interface and their potential methane contribution to the atmosphere is under investigation. Another surprising observation is the absence of major seafloor features like pockmarks in the gas release area. The reasons for the seafloor methane release in the Landsort deep are still not entirely clear, but our preliminary acoustic and sedimentological data suggest that bottom currents may have acted to facilitate the accumulation of organic-rich sediments in a thick drift deposit during the Holocene and the modern warm period (latest 100 years). Our data further suggest that the high sedimentation rate in the drift deposit continuously supplies fresh organic matter that is quickly buried below a thin sulphate reduction zone, fueling vigorous methanogenesis and abundant methane formation. Similar methane release sites might be discovered in other known large drift deposits in the Baltic Sea.

Many individuals who refuse COVID-19 vaccination have concerns about long-term side effects. Here, we report findings on self-reported symptoms from a Danish survey- and register study. The study included 34,868 vaccinated primary course recipients, 95.8% of whom received mRNA vaccines, and 1,568 unvaccinated individuals. Participants had no known history of SARS-CoV-2 infection. Using g-computation on logistic regression, risk differences (RDs) for symptoms between vaccinated and unvaccinated persons were estimated with adjustments for possible confounders. Within six weeks after vaccination, higher risks were observed for physical exhaustion (RD 4.9%, 95% CI 1.1% to 8.4%), fever or chills (RD 4.4%, 95% CI 2.1% to 6.7%), and muscle/joint pain (RD 7.0%, 95% CI 3.1% to 10.7%), compared to unvaccinated individuals. Beyond twenty-six weeks, risks were higher among the vaccinated for sleeping problems (RD 3.0, 95% 0.2 to 5.8), fever or chills (RD 2.0, 95% CI 0.4 to 3.6), reduced/altered taste (RD 1.2, 95% CI 0.2 to 2.3) and shortness of breath (RD 2.6, 95% CI 0.9 to 4.0). However, when examining pre-omicron responses only, the difference for reduced/altered taste was significant. As expected, the risk of experiencing physical exhaustion, fever or chills, and muscle/joint pain was higher among persons who responded within six weeks of completing the primary course. No significant differences were observed for the 7-25-week period after vaccination. Associations for the period beyond 26 weeks must be interpreted with caution and in the context of undetected SARS-CoV-2 infection, wide confidence intervals, and multiple testing. Overall, we observe no concerning signs of long-term self-reported physical, cognitive, or fatigue symptoms after vaccination.

Qi Rui See,1,* Kausar Raheel,1,* Iain Duncan,1,* Nazanin Biabani,1,* Irene Di Giulio,2 Andrea Romigi,3 Veena Kumari,1,4 David O’Regan,1,2,5 Scott Cairney,6 Daniele Urso,7,8 K Ray Chaudhuri,8 Valentina Gnoni,1,7 Panagis Drakatos,1,2,5,* Ivana Rosenzweig1,5,* 1Sleep and Brain Plasticity Centre, Department of Neuroimaging, Institute of Psychiatry, Psychology and Neuroscience (IoPPN), King’s College London, London, U.K; 2School of Basic and Medical Biosciences, Faculty of Life Science and Medicine, King’s College London, London, U.K; 3IRCCS Neuromed Istituto Neurologico Mediterraneo Pozzilli (IS), Pozzilli, Italy; 4Centre for Cognitive Neuroscience, College of Health, Medicine and Life Sciences, Brunel University London, Uxbridge, U.K; 5Sleep Disorders Centre, Guy’s and St Thomas’ NHS Foundation Trust, London, U.K; 6Department of Psychology, University of York and York Biomedical Research Institute, University of York, York, U.K; 7Center for Neurodegenerative Diseases and the Aging Brain, Department of Clinical Research in Neurology, University of Bari ‘Aldo Moro’, “Pia Fondazione Cardinale G. Panico”, Tricase, Lecce, Italy; 8Movement Disorders Unit, King’s College Hospital and Department of Clinical and Basic Neurosciences, Institute of Psychiatry, Psychology & Neuroscience and Parkinson Foundation Centre of Excellence, King’s College London, London, U.K*These authors contributed equally to this workCorrespondence: Ivana Rosenzweig, Sleep and Brain Plasticity Centre, Department of Neuroimaging, Box 089, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, De Crespigny Park, London, SE5 8AF, UK, Email ivana.1.rosenzweig@kcl.ac.ukBackground: Speech graph analysis (SGA) of dreams has recently shown promise as an objective and language-invariant diagnostic tool that can aid neuropsychiatric diagnosis. Whilst the notion that dreaming mentations reflect distinct physio

Many individuals who refuse COVID-19 vaccination have concerns about long-term side effects. Here, we report findings on self-reported symptoms from a Danish survey- and register study. The study included 34,868 vaccinated primary course recipients, 95.8% of whom received mRNA vaccines, and 1,568 unvaccinated individuals. Participants had no known history of SARS-CoV-2 infection. Using g-computation on logistic regression, risk differences (RDs) for symptoms between vaccinated and unvaccinated persons were estimated with adjustments for possible confounders. Within six weeks after vaccination, higher risks were observed for physical exhaustion (RD 4.9%, 95% CI 1.1% to 8.4%), fever or chills (RD 4.4%, 95% CI 2.1% to 6.7%), and muscle/joint pain (RD 7.0%, 95% CI 3.1% to 10.7%), compared to unvaccinated individuals. Beyond twenty-six weeks, risks were higher among the vaccinated for sleeping problems (RD 3.0, 95% 0.2 to 5.8), fever or chills (RD 2.0, 95% CI 0.4 to 3.6), reduced/altered taste (RD 1.2, 95% CI 0.2 to 2.3) and shortness of breath (RD 2.6, 95% CI 0.9 to 4.0). However, when examining pre-omicron responses only, the difference for reduced/altered taste was significant. As expected, the risk of experiencing physical exhaustion, fever or chills, and muscle/joint pain was higher among persons who responded within six weeks of completing the primary course. No significant differences were observed for the 7-25-week period after vaccination. Associations for the period beyond 26 weeks must be interpreted with caution and in the context of undetected SARS-CoV-2 infection, wide confidence intervals, and multiple testing. Overall, we observe no concerning signs of long-term self-reported physical, cognitive, or fatigue symptoms after vaccination.

Depto. de Química en Ciencias Farmacéuticas, Fac. de Farmacia, TRUE, pub

Depto. de Química en Ciencias Farmacéuticas, Fac. de Farmacia, TRUE, pub

Depto. de Química en Ciencias Farmacéuticas, Fac. de Farmacia, TRUE, pub

Surrogacy, a globally researched topic, often side lines the voices of surrogate mothers, primarily from developing countries and is often viewed through a Western lens. Grounded in social constructionism, this study delved into Indian surrogate mothers' experiences via narrative inquiry. 24 surrogate mothers in the state of Karnataka, India, were selected through purposive sampling, to capture a varied range of narratives, allowing the participants to express their distinctive experiences of surrogacy. Using Fraser's seven phases of narrative inquiry, the findings unveiled the complexities and nuances of these experiences for the surrogates. It further highlighted their sense of agency and empowerment while acknowledging the struggles (such as experiences of obstetric violence) within the socio-cultural and political context domestically. It also questioned the overarching systemic issues created by the fertility industry and what it implied for developing countries offering fertility services. This research contributes to the existing literature on surrogacy and offers insights that will inform policy and practice to safeguard the rights and wellbeing of Indian surrogate mothers. It emphasises the need for hearing and acknowledging surrogate mothers' voices in shaping future policies and ethical frameworks surrounding surrogacy., Thesis (PhD Doctorate), Doctor of Philosophy (PhD), School of Health Sci & Soc Wrk, Griffith Health, Full Text

Although humans engaged in face-to-face conversation simultaneously communicate both verbally and non-verbally, methods for joint and unified synthesis of speech audio and co-speech 3D gesture motion from text are a new and emerging field. These technologies hold great promise for more human-like, efficient, expressive, and robust synthetic communication, but are currently held back by the lack of suitably large datasets, as existing methods are trained on parallel data from all constituent modalities. Inspired by student-teacher methods, we propose a straightforward solution to the data shortage, by simply synthesising additional training material. Specifically, we use unimodal synthesis models trained on large datasets to create multimodal (but synthetic) parallel training data, and then pre-train a joint synthesis model on that material. In addition, we propose a new synthesis architecture that adds better and more controllable prosody modelling to the state-of-the-art method in the field. Our results confirm that pre-training on large amounts of synthetic data improves the quality of both the speech and the motion synthesised by the multimodal model, with the proposed architecture yielding further benefits when pre-trained on the synthetic data. See https://shivammehta25.github.io/MAGI/ for example output., Comment: 13+1 pages, 2 figures, accepted at the Human Motion Generation workshop (HuMoGen) at CVPR 2024

In recent years we have seen a growing recognition by educators of the value of Virtual Exchange (VE) in the context of Internationalisation at Home (IaH) (de Wit, 2016). Research has shown the benefits of VE for language learners, as it emphasises the development of transferable skills (European Commission, 2021; Godwin-Jones, 2020). In this paper, we report on the development of Erasmus Speaks, an innovative, transnational Virtual Exchange Project offered at a third-level research institution in Ireland in autumn 2020. This new element replaced the Erasmus study-abroad period for over 600 European language students in response to Covid-19 pandemic travel restrictions. The high completion rate of the project, reflected in the large number of Transnational Erasmus+ Project (TEP) digital badges awarded to participants, shows that this task-based project met its objectives in relation to language learning, intercultural awareness, digital literacy and transferable skills.

The extent and seasonality of Arctic sea ice during the Last Interglacial (129,000 to 115,000 years before present) is poorly known. Sediment-based reconstructions have suggested extensive ice cover in summer, while climate model outputs indicate year-round conditions in the Arctic Ocean ranging from ice free to fully ice covered. Here we use microfossil records from across the central Arctic Ocean to show that sea-ice extent was substantially reduced and summers were probably ice free. The evidence comes from high abundances of the subpolar planktic foraminifera Turborotalita quinqueloba in five newly analysed cores. The northern occurrence of this species is incompatible with perennial sea ice, which would be associated with a thick, low-salinity surface water. Instead, T. quinqueloba's ecological preference implies largely ice-free surface waters with seasonally elevated levels of primary productivity. In the modern ocean, this species thrives in the Fram Strait-Barents Sea 'Arctic-Atlantic gateway' region, implying that the necessary Atlantic Ocean-sourced water masses shoaled towards the surface during the Last Interglacial. This process reflects the ongoing Atlantification of the Arctic Ocean, currently restricted to the Eurasian Basin. Our results establish the Last Interglacial as a prime analogue for studying a seasonally ice-free Arctic Ocean, expected to occur this century. The warm Last Interglacial led to a seasonally ice-free Arctic Ocean and a transformation to Atlantic conditions, according to planktic foraminifera records from central Arctic Ocean sediment cores.

Using a conceptual model, we examine how hydraulically controlled exchange flows in silled fjords affect the relationship between the basal glacier melt and the features of warm intermediate Atlantic Water (AW) outside the fjords. We show that an exchange flow can be forced to transit into the hydraulic regime if the AW interface height decreases, the AW temperature increases, or the production of glacially modified water is boosted by subglacial discharge. In the hydraulic regime, the heat transport across the sill becomes a rate-limiting factor for the basal melt, which is suppressed. An interplay between processes near the ice-ocean boundary and the hydraulically controlled exchange flow determines the melt dynamics, and the sensitivity of the basal melt to changes in the AW temperature is reduced. The model results are discussed in relation to observations from the Petermann, Ryder, and 79∘ N glaciers in northern Greenland.

According to climate models, the Lincoln Sea, bordering northern Greenland and Canada, will be the final stronghold of perennial Arctic sea-ice in a warming climate. However, recent observations of prolonged periods of open water raise concerns regarding its long-term stability. Modelling studies suggest a transition from perennial to seasonal sea-ice during the Early Holocene, a period of elevated global temperatures around 10,000 years ago. Here we show marine proxy evidence for the disappearance of perennial sea-ice in the southern Lincoln Sea during the Early Holocene, which suggests a widespread transition to seasonal sea-ice in the Arctic Ocean. Seasonal sea-ice conditions were tightly coupled to regional atmospheric temperatures. In light of anthropogenic warming and Arctic amplification our results suggest an imminent transition to seasonal sea-ice in the southern Lincoln Sea, even if the global temperature rise is kept below a threshold of 2 °C compared to pre-industrial (1850–1900).

The unprecedented progress in ensemble hydro-meteorological modelling and forecasting on a range of temporal and spatial scales, raises a variety of new challenges which formed the theme of the Joint Virtual Workshop, ‘Connecting global to local hydrological modelling and forecasting: challenges and scientific advances’. Held from 29 June to 1 July 2021, this workshop was co-organised by the European Centre for Medium-Range Weather Forecasts (ECMWF), the Copernicus Emergency Management (CEMS) and Climate Change (C3S) Services, the Hydrological Ensemble Prediction EXperiment (HEPEX), and the Global Flood Partnership (GFP). This article aims to summarise the state-of-the-art presented at the workshop and provide an early career perspective. Recent advances in hydrological modelling and forecasting, reflections on the use of forecasts for decision-making across scales, and means to minimise new barriers to communication in the virtual format are also discussed. Thematic foci of the workshop included hydrological model development and skill assessment, uncertainty communication, forecasts for early action, co-production of services and incorporation of local knowledge, Earth observation, and data assimilation. Connecting hydrological services to societal needs and local decision-making through effective communication, capacity-building and co-production was identified as critical. Multidisciplinary collaborations emerged as crucial to effectively bring newly developed tools to practice.

Despite extensive chronological studies, the relationship between the age and sub-seafloor depth of Arctic Ocean sediments remains ambiguous. This prevents confident identification of paleoceanographic changes in the Arctic during the Quaternary. Currently, age-depth models derived from uranium-series decay in Arctic sediments diverge by hundreds of thousands of years compared to those built on known evolutionary appearances and extinctions of calcareous nannoplankton, a group of globally valuable age-markers. Here we report on highresolution biostratigraphic analysis of late Quaternary sediments in six cores from the central Arctic Ocean (CAO). We applied paired light microscope (LM) and scanning electron microscope (SEM) imaging to improve nannofossil diagnosis. We argue that low abundances and poor preservation have led to misidentification of the true stratigraphic depth of the critical Pleistocene nannofossil bio-events that have underpinned age models for many Arctic sedimentary records for decades. The revised calcareous nannofossil biochronology provides a radically different geochronological framework for CAO sediments - indicating that what had previously been identified as Marine Isotope Stage (MIS) 7 (191-243 ka) in many sedimentary records is older than MIS 12 (424-478 ka). Furthermore, it suggests that previously inferred sub-stages of MIS 5 could represent full interglacial periods rather than interstadials. The results help reconcile the different dating approaches and provide a transformative step towards resolving the disparity in Quaternary Arctic age-depth models, bringing us one step closer to accurate paleoceanographic reconstructions based on sediment cores.

Background: In an earlier interview-based study the authors identified that learners experience one or more of eight explicit perceptual responses during the active phase of simulation-based training (SBT) comprising a sense: of belonging to instructor and group, of being under surveillance, of having autonomy and responsibility for patient management, of realism, of an understanding of the scenario in context, of conscious mental effort, of control of attention, and of engagement with task. These were adapted into a ten-item questionnaire: the Simulation Based Training Quality Assurance Tool (SBT-QA10) to allow monitoring of modifiable factors that may impact upon learners’ experiences. This study assessed the construct validity evidence of the interpretation of the results when using SBT-QAT10. Materials and methods: Recently graduated doctors and nurses participating in a SBT course on the topic of the deteriorating patient completed the SBT-QAT10 immediately following their participation in the scenarios. The primary outcome measure was internal consistency of the questionnaire items and their correlation to learners’ satisfaction scores. A secondary outcome measure compared the impact of allocation to active versus observer role. Results: A total of 349 questionnaires were returned by 96 course learners. The median of the total score for the ten perception items (TPS) was 39 (out of 50), with no significant difference between the scenarios. We identified fair and positive correlations between nine of the 10 items and the SBT-QA10-TPS, the exception being “mental effort”. Compared to observers, active learners reported significantly more positive perceptions related to belonging to the team and interaction with the instructor, their sense of acting independently, and being focused. The questionnaire items were poorly correlated with the two measures of global satisfaction. Conclusion: Except for the item for mental effort, the QA10-TPS measures learners’ expe

Post-acute sick leave is an underexplored indicator of the societal burden of SARS-CoV-2. Here, we report findings about self-reported sick leave and risk factors thereof from a hybrid survey and register study, which include 37,482 RT-PCR confirmed SARS-CoV-2 cases and 51,336 test-negative controls who were tested during the index- and alpha-dominant waves. We observe that an additional 33 individuals per 1000 took substantial sick leave following acute infection compared to persons with no known history of infection, where substantial sick leave is defined as >1 month of sick leave within the period 1–9 months after the RT-PCR test date. Being female, 50–65 years, or having certain pre-existing health conditions such as obesity, chronic lung diseases, and fibromyalgia each increase risk for taking substantial sick leave. Altogether, these results may help motivate improved diagnostic and treatment options for persons living with post-Covid conditions.

The analysis of 87Sr/86Sr isotope ratios in human and nonhuman tooth enamel is used worldwide for archaeological and forensic purposes to establish if an individual is likely to have grown up in the area from which their remains were excavated. The English Peak District has produced an unusually high proportion of archaeological humans who, based on Sr isotope ratios, appear to have come from elsewhere. We have used modern plant samples from the Peak to show that the current understanding of Sr isotope ratios for this area is incomplete – we found many plant samples growing on gritstone sediments had higher Sr values than would be expected based on the current literature. In addition we demonstrated that the taxonomy of the plant does not appear to affect the Sr isotope values (we also found that mycorrhizal type did not determine Sr isotope values in these plants), rather it is the substrate on which it is growing that is important. In terms of human movement, our work suggests it is likely that many archaeological individuals found in the Peak District are indeed local, rather than migrants. It is also possible that the expansion of blanket peat in the Peak has over time reduced the amount of Sr entering the food chain from mineral soils, reducing the radiogenic Sr isotope values in more recent teeth. While our case study is the Peak District, our findings have implications for anomalously high archaeological 87Sr/86Sr isotope values in other upland regions with similar geologies and blanket peats.

University patents are a critical tool for firms seeking to gain information from universities. However, the potential of this tool is frequently constrained by ineffective communication and commercialisation strategies. University patents must be successfully advertised and disseminated to third parties to have an impact outside academia. In this scenario, inventors, universities, and technology transfer offices (TTOs) (which manage their patent portfolios) are encouraged to increase their efforts to promote and communicate patents. This article uses signalling theory to examine the key features of university patents that influence the investment or acquisition decisions of entrepreneurs and investors. Our findings can help TTOs, inventors, and universities strengthen their patent communication and commercialisation strategies, and also help third parties secure more successful university patents.

This paper investigates the impact of four drivers of Governance diversity, namely gender, tenure, age, and educational attainment, on strategic direction and variation. It then incorporates corporate financial results as a moderating variable, testing how it impacts the links between board diversity and strategic variation. Strategic variation or change is assessed based on measuring deviation from past strategies. Our sample consists of 5011 firm-year observations from 930 firms in the United States between 2010 and 2018. The findings indicate that the four drivers of Governance diversity show a positive relationship between the corporate board of directors and strategic variation/change. However, the strength of the effect depends on overall firm results/performance.

University patents are a critical tool for firms seeking to gain information from universities. However, the potential of this tool is frequently constrained by ineffective communication and commercialisation strategies. University patents must be successfully advertised and disseminated to third parties to have an impact outside academia. In this scenario, inventors, universities, and technology transfer offices (TTOs) (which manage their patent portfolios) are encouraged to increase their efforts to promote and communicate patents. This article uses signalling theory to examine the key features of university patents that influence the investment or acquisition decisions of entrepreneurs and investors. Our findings can help TTOs, inventors, and universities strengthen their patent communication and commercialisation strategies, and also help third parties secure more successful university patents.

This paper investigates the impact of four drivers of Governance diversity, namely gender, tenure, age, and educational attainment, on strategic direction and variation. It then incorporates corporate financial results as a moderating variable, testing how it impacts the links between board diversity and strategic variation. Strategic variation or change is assessed based on measuring deviation from past strategies. Our sample consists of 5011 firm-year observations from 930 firms in the United States between 2010 and 2018. The findings indicate that the four drivers of Governance diversity show a positive relationship between the corporate board of directors and strategic variation/change. However, the strength of the effect depends on overall firm results/performance.

Two key questions in cardiac image analysis are to assess the anatomy and motion of the heart from images; and to understand how they are associated with non-imaging clinical factors such as gender, age and diseases. While the first question can often be addressed by image segmentation and motion tracking algorithms, our capability to model and to answer the second question is still limited. In this work, we propose a novel conditional generative model to describe the 4D spatio-temporal anatomy of the heart and its interaction with non-imaging clinical factors. The clinical factors are integrated as the conditions of the generative modelling, which allows us to investigate how these factors influence the cardiac anatomy. We evaluate the model performance in mainly two tasks, anatomical sequence completion and sequence generation. The model achieves a high performance in anatomical sequence completion, comparable to or outperforming other state-of-the-art generative models. In terms of sequence generation, given clinical conditions, the model can generate realistic synthetic 4D sequential anatomies that share similar distributions with the real data., Comment: Accepted by IEEE Transactions on Medical Imaging

Background and Objective: Numerical simulations of blood flow are a valuable tool to investigate the pathophysiology of ascending thoratic aortic aneurysms (ATAA). To accurately reproduce in vivo hemodynamics, computational fluid dynamics (CFD) models must employ realistic inflow boundary conditions (BCs). However, the limited availability of in vivo velocity measurements, still makes researchers resort to idealized BCs. The aim of this study was to generate and thoroughly characterize a large dataset of synthetic 4D aortic velocity profiles sampled on a 2D cross-section along the ascending aorta with features similar to clinical cohorts of patients with ATAA. Methods: Time-resolved 3D phase contrast magnetic resonance (4D flow MRI) scans of 30 subjects with ATAA were processed through in-house code to extract anatomically consistent cross-sectional planes along the ascending aorta, ensuring spatial alignment among all planes and interpolating all velocity fields to a reference configuration. Velocity profiles of the clinical cohort were extensively characterized by computing flow morphology descriptors of both spatial and temporal features. By exploiting principal component analysis (PCA), a statistical shape model (SSM) of 4D aortic velocity profiles was built and a dataset of 437 synthetic cases with realistic properties was generated. Results: Comparison between clinical and synthetic datasets showed that the synthetic data presented similar characteristics as the clinical population in terms of key morphological parameters. The average velocity profile qualitatively resembled a parabolic-shaped profile, but was quantitatively characterized by more complex flow patterns which an idealized profile would not replicate. Statistically significant correlations were found between PCA principal modes of variation and flow descriptors. Conclusions: We built a data-driven generative model of 4D aortic inlet velocity profiles, suitable to be used in computational studi, Medical Instruments & Bio-Inspired Technology

The unprecedented progress in ensemble hydro-meteorological modelling and forecasting on a range of temporal and spatial scales, raises a variety of new challenges which formed the theme of the Joint Virtual Workshop, ‘Connecting global to local hydrological modelling and forecasting: challenges and scientific advances’. Held from 29 June to 1 July 2021, this workshop was co-organised by the European Centre for Medium-Range Weather Forecasts (ECMWF), the Copernicus Emergency Management (CEMS) and Climate Change (C3S) Services, the Hydrological Ensemble Prediction EXperiment (HEPEX), and the Global Flood Partnership (GFP). This article aims to summarise the state-of-the-art presented at the workshop and provide an early career perspective. Recent advances in hydrological modelling and forecasting, reflections on the use of forecasts for decision-making across scales, and means to minimise new barriers to communication in the virtual format are also discussed. Thematic foci of the workshop included hydrological model development and skill assessment, uncertainty communication, forecasts for early action, co-production of services and incorporation of local knowledge, Earth observation, and data assimilation. Connecting hydrological services to societal needs and local decision-making through effective communication, capacity-building and co-production was identified as critical. Multidisciplinary collaborations emerged as crucial to effectively bring newly developed tools to practice.

How can we interpret and retrieve medical evidence to support clinical decisions? Clinical trial reports (CTR) amassed over the years contain indispensable information for the development of personalized medicine. However, it is practically infeasible to manually inspect over 400,000+ clinical trial reports in order to find the best evidence for experimental treatments. Natural Language Inference (NLI) offers a potential solution to this problem, by allowing the scalable computation of textual entailment. However, existing NLI models perform poorly on biomedical corpora, and previously published datasets fail to capture the full complexity of inference over CTRs. In this work, we present a novel resource to advance research on NLI for reasoning on CTRs. The resource includes two main tasks. Firstly, to determine the inference relation between a natural language statement, and a CTR. Secondly, to retrieve supporting facts to justify the predicted relation. We provide NLI4CT, a corpus of 2400 statements and CTRs, annotated for these tasks. Baselines on this corpus expose the limitations of existing NLI models, with 6 state-of-the-art NLI models achieving a maximum F1 score of 0.627. To the best of our knowledge, we are the first to design a task that covers the interpretation of full CTRs. To encourage further work on this challenging dataset, we make the corpus, competition leaderboard, website and code to replicate the baseline experiments available at: https://github.com/ai-systems/nli4ct, Comment: EMNLP 2023 Camera-ready, 15 pages

This paper describes the results of SemEval 2023 task 7 -- Multi-Evidence Natural Language Inference for Clinical Trial Data (NLI4CT) -- consisting of 2 tasks, a Natural Language Inference (NLI) task, and an evidence selection task on clinical trial data. The proposed challenges require multi-hop biomedical and numerical reasoning, which are of significant importance to the development of systems capable of large-scale interpretation and retrieval of medical evidence, to provide personalized evidence-based care. Task 1, the entailment task, received 643 submissions from 40 participants, and Task 2, the evidence selection task, received 364 submissions from 23 participants. The tasks are challenging, with the majority of submitted systems failing to significantly outperform the majority class baseline on the entailment task, and we observe significantly better performance on the evidence selection task than on the entailment task. Increasing the number of model parameters leads to a direct increase in performance, far more significant than the effect of biomedical pre-training. Future works could explore the limitations of large models for generalization and numerical inference, and investigate methods to augment clinical datasets to allow for more rigorous testing and to facilitate fine-tuning. We envisage that the dataset, models, and results of this task will be useful to the biomedical NLI and evidence retrieval communities. The dataset, competition leaderboard, and website are publicly available.

Independent Vector Analysis (IVA) is a popular extension of Independent Component Analysis (ICA) for joint separation of a set of instantaneous linear mixtures, with a direct application in frequency-domain speaker separation or extraction. The mixtures are parameterized by mixing matrices, one matrix per mixture. This means that the IVA mixing model does not account for any relationships between parameters across the mixtures/frequencies. The separation proceeds jointly only through the source model, where statistical dependencies of sources across the mixtures are taken into account. In this paper, we propose a mixing model for joint blind source extraction where the mixing model parameters are linked across the frequencies. This is achieved by constraining the set of feasible parameters to the manifold of half-length separating filters, which has a clear interpretation and application in frequency-domain speaker extraction.

BACKGROUND: Post COVID-19 condition (PCC) can lead to considerable morbidity, including prolonged sick-leave. Identifying risk groups is important for informing interventions. We investigated heterogeneity in the effect of SARS-CoV-2 infection on long-term sick-leave and identified subgroups at higher risk.METHODS: We conducted a hybrid survey and register-based retrospective cohort study of Danish residents who tested positive for SARS-CoV-2 between November 2020 and February 2021 and a control group who tested negative, with no known history of SARS-CoV-2. We estimated the causal risk difference (RD) of long-term sick-leave due to PCC and used the causal forest method to identify individual-level heterogeneity in the effect of infection on sick-leave. Sick-leave was defined as >4 weeks of full-time sick-leave from 4 weeks to 9 months after the test.RESULTS: Here, in a cohort of 88,818 individuals, including 37,482 with a confirmed SARS-CoV-2 infection, the RD of long-term sick-leave is 3.3% (95% CI 3.1% to 3.6%). We observe a high degree of effect heterogeneity, with conditional RDs ranging from -3.4% to 13.7%. Age, high BMI, depression, and sex are the most important variables explaining heterogeneity. Among three-way interactions considered, females with high BMI and depression and persons aged 36-45 years with high BMI and depression have an absolute increase in risk of long-term sick-leave above 10%.CONCLUSIONS: Our study supports significant individual-level heterogeneity in the effect of SARS-CoV-2 infection on long-term sick-leave, with age, sex, high BMI, and depression identified as key factors. Efforts to curb the PCC burden should consider multimorbidity and individual-level risk.

Background: The complex genetics underlying human cardiac disease is evidenced by its heterogenous manifestation, multigenic basis, and sporadic occurrence. These features have hampered disease modeling and mechanistic understanding. Here, we show that 2 structural cardiac diseases, left ventricular noncompaction (LVNC) and bicuspid aortic valve, can be caused by a set of inherited heterozygous gene mutations affecting the NOTCH ligand regulator MIB1 (MINDBOMB1) and cosegregating genes. Methods: We used CRISPR-Cas9 gene editing to generate mice harboring a nonsense or a missense MIB1 mutation that are both found in LVNC families. We also generated mice separately carrying these MIB1 mutations plus 5 additional cosegregating variants in the ASXL3, APCDD1, TMX3, CEP192, and BCL7A genes identified in these LVNC families by whole exome sequencing. Histological, developmental, and functional analyses of these mouse models were carried out by echocardiography and cardiac magnetic resonance imaging, together with gene expression profiling by RNA sequencing of both selected engineered mouse models and human induced pluripotent stem cell-derived cardiomyocytes. Potential biochemical interactions were assayed in vitro by coimmunoprecipitation and Western blot. Results: Mice homozygous for the MIB1 nonsense mutation did not survive, and the mutation caused LVNC only in heteroallelic combination with a conditional allele inactivated in the myocardium. The heterozygous MIB1 missense allele leads to bicuspid aortic valve in a NOTCH-sensitized genetic background. These data suggest that development of LVNC is influenced by genetic modifiers present in affected families, whereas valve defects are highly sensitive to NOTCH haploinsufficiency. Whole exome sequencing of LVNC families revealed single-nucleotide gene variants of ASXL3, APCDD1, TMX3, CEP192, and BCL7A cosegregating with the MIB1 mutations and LVNC. In experiments with mice harboring the orthologous variants on the cor

Background: In an earlier interview-based study the authors identified that learners experience one or more of eight explicit perceptual responses during the active phase of simulation-based training (SBT) comprising a sense: of belonging to instructor and group, of being under surveillance, of having autonomy and responsibility for patient management, of realism, of an understanding of the scenario in context, of conscious mental effort, of control of attention, and of engagement with task. These were adapted into a ten-item questionnaire: the Simulation Based Training Quality Assurance Tool (SBT-QA10) to allow monitoring of modifiable factors that may impact upon learners’ experiences. This study assessed the construct validity evidence of the interpretation of the results when using SBT-QAT10. Materials and methods: Recently graduated doctors and nurses participating in a SBT course on the topic of the deteriorating patient completed the SBT-QAT10 immediately following their participation in the scenarios. The primary outcome measure was internal consistency of the questionnaire items and their correlation to learners’ satisfaction scores. A secondary outcome measure compared the impact of allocation to active versus observer role. Results: A total of 349 questionnaires were returned by 96 course learners. The median of the total score for the ten perception items (TPS) was 39 (out of 50), with no significant difference between the scenarios. We identified fair and positive correlations between nine of the 10 items and the SBT-QA10-TPS, the exception being “mental effort”. Compared to observers, active learners reported significantly more positive perceptions related to belonging to the team and interaction with the instructor, their sense of acting independently, and being focused. The questionnaire items were poorly correlated with the two measures of global satisfaction. Conclusion: Except for the item for mental effort, the QA10-TPS measures learners’ expe

Post-acute sick leave is an underexplored indicator of the societal burden of SARS-CoV-2. Here, we report findings about self-reported sick leave and risk factors thereof from a hybrid survey and register study, which include 37,482 RT-PCR confirmed SARS-CoV-2 cases and 51,336 test-negative controls who were tested during the index- and alpha-dominant waves. We observe that an additional 33 individuals per 1000 took substantial sick leave following acute infection compared to persons with no known history of infection, where substantial sick leave is defined as >1 month of sick leave within the period 1–9 months after the RT-PCR test date. Being female, 50–65 years, or having certain pre-existing health conditions such as obesity, chronic lung diseases, and fibromyalgia each increase risk for taking substantial sick leave. Altogether, these results may help motivate improved diagnostic and treatment options for persons living with post-Covid conditions.

BACKGROUND: Post COVID-19 condition (PCC) can lead to considerable morbidity, including prolonged sick-leave. Identifying risk groups is important for informing interventions. We investigated heterogeneity in the effect of SARS-CoV-2 infection on long-term sick-leave and identified subgroups at higher risk.METHODS: We conducted a hybrid survey and register-based retrospective cohort study of Danish residents who tested positive for SARS-CoV-2 between November 2020 and February 2021 and a control group who tested negative, with no known history of SARS-CoV-2. We estimated the causal risk difference (RD) of long-term sick-leave due to PCC and used the causal forest method to identify individual-level heterogeneity in the effect of infection on sick-leave. Sick-leave was defined as >4 weeks of full-time sick-leave from 4 weeks to 9 months after the test.RESULTS: Here, in a cohort of 88,818 individuals, including 37,482 with a confirmed SARS-CoV-2 infection, the RD of long-term sick-leave is 3.3% (95% CI 3.1% to 3.6%). We observe a high degree of effect heterogeneity, with conditional RDs ranging from -3.4% to 13.7%. Age, high BMI, depression, and sex are the most important variables explaining heterogeneity. Among three-way interactions considered, females with high BMI and depression and persons aged 36-45 years with high BMI and depression have an absolute increase in risk of long-term sick-leave above 10%.CONCLUSIONS: Our study supports significant individual-level heterogeneity in the effect of SARS-CoV-2 infection on long-term sick-leave, with age, sex, high BMI, and depression identified as key factors. Efforts to curb the PCC burden should consider multimorbidity and individual-level risk.

3D motion estimation from cine cardiac magnetic resonance (CMR) images is important for the assessment of cardiac function and the diagnosis of cardiovascular diseases. Current state-of-the art methods focus on estimating dense pixel-/voxel-wise motion fields in image space, which ignores the fact that motion estimation is only relevant and useful within the anatomical objects of interest, e.g., the heart. In this work, we model the heart as a 3D mesh consisting of epi- and endocardial surfaces. We propose a novel learning framework, DeepMesh, which propagates a template heart mesh to a subject space and estimates the 3D motion of the heart mesh from CMR images for individual subjects. In DeepMesh, the heart mesh of the end-diastolic frame of an individual subject is first reconstructed from the template mesh. Mesh-based 3D motion fields with respect to the end-diastolic frame are then estimated from 2D short- and long-axis CMR images. By developing a differentiable mesh-to-image rasterizer, DeepMesh is able to leverage 2D shape information from multiple anatomical views for 3D mesh reconstruction and mesh motion estimation. The proposed method estimates vertex-wise displacement and thus maintains vertex correspondences between time frames, which is important for the quantitative assessment of cardiac function across different subjects and populations. We evaluate DeepMesh on CMR images acquired from the UK Biobank. We focus on 3D motion estimation of the left ventricle in this work. Experimental results show that the proposed method quantitatively and qualitatively outperforms other image-based and mesh-based cardiac motion tracking methods.

Since the buildings and construction sector is one of the main areas responsible for energy consumption and emissions, focusing on their refurbishment and promoting actions in this direction will be helpful to achieve an EU Agenda objective of making Europe climate-neutral by 2050. One step towards the renovation action is the exploitation of digital tools into a BIM framework. The scope of the research contained in this paper is to improve the management of information throughout the different stages of the renovation process, allowing an interoperable exchange of data among the involved stakeholders; the development of an innovative BIM-based toolkit is the answer to the research question. The research and results obtained related with the development of an interoperable BIM-based toolkit for efficient renovation in buildings in the framework of the European research project BIM4EEB. Specifically, the developed BIM management system allows the exchange of the data among the different tools, using open interoperable formats (as IFC) and linked data, in a Common Data Environment, to be used by the different stakeholders. Additionally, the developed tools allow the stakeholders to manage different stages of the renovation process, facilitating efficiencies in terms of time reduction and improving the resulting quality. The validity of each tool with respect to existing practices is demonstrated here, and the strengths and weaknesses of the proposed tools are described in the workflow detailing issues such as interoperability, collaboration, integration of different solutions, and time consuming existing survey processes.

Organizational agility, that is the ability to anticipate or respond quickly to external changes, is essential to survive and compete in today's turbulent landscape, characterized by technological advancements and digitalization. Research on capabilities that enable firms to be agile in the so called VUCA environments, is still nascent. Hence, it is important to explore the antecedents of firm agility and to identify the factors enabling them to better compete. Even more so in the case of SMEs, as they are more vulnerable in hypercompetitive business environments and at the same time agility has been less studied in this context. Focusing on SMEs, the study investigates three antecedents of agility, namely digital technologies capability, relational capability and innovation capability, and the effects of agility on three outcomes, namely financial performance, product, and process innovation. Our findings indicate that these capabilities contribute to build organizational agility in SMEs and that, in turn, agility has a positive impact on performance, thus confirming that agility contributes to the success of SMEs and that digital technologies play a central role in this process. Thus, it is of strategic importance for SMEs to increase their efforts to develop these capabilities to build enduring businesses. They should nurture a relational and innovative culture, as well as transform their business culture starting from digital technologies.

Organizational agility, that is the ability to anticipate or respond quickly to external changes, is essential to survive and compete in today's turbulent landscape, characterized by technological advancements and digitalization. Research on capabilities that enable firms to be agile in the so called VUCA environments, is still nascent. Hence, it is important to explore the antecedents of firm agility and to identify the factors enabling them to better compete. Even more so in the case of SMEs, as they are more vulnerable in hypercompetitive business environments and at the same time agility has been less studied in this context. Focusing on SMEs, the study investigates three antecedents of agility, namely digital technologies capability, relational capability and innovation capability, and the effects of agility on three outcomes, namely financial performance, product, and process innovation. Our findings indicate that these capabilities contribute to build organizational agility in SMEs and that, in turn, agility has a positive impact on performance, thus confirming that agility contributes to the success of SMEs and that digital technologies play a central role in this process. Thus, it is of strategic importance for SMEs to increase their efforts to develop these capabilities to build enduring businesses. They should nurture a relational and innovative culture, as well as transform their business culture starting from digital technologies.