13463 research outputs found
Sort by
On the usage of artificial intelligence in leprosy care: A systematic literature review
Full text linkLeprosy, or Hansen’s disease, is a Neglected Tropical Disease (NTD) caused by Mycobacterium leprae that mainly affects the skin and peripheral nerves, causing neuropathy to varying degrees. It can result in physical disabilities and functional loss and is particularly prevalent amongst the most vulnerable populations in tropical and subtropical regions worldwide. The persistent stigma and social exclusion associated with leprosy complicate eradication efforts exacerbate the wider challenges faced by NTDs in sourcing the necessary resources and attention for control and elimination. The introduction of Multidrug Therapy (MDT) significantly lowers the global disease burden. Despite this breakthrough in the treatment of leprosy, over 200,000 new leprosy cases are reported annually across more than 120 countries, emphasizing the need for ongoing detection and management efforts. Artificial Intelligence (AI) has the potential to transform leprosy care by accelerating early detection, improving accurate diagnosis, and enabling
predictive modeling to improve the quality for those affected. The potential of AI to provide information to assist healthcare professionals in interventions that reduce the risk of disability, and consequently stigma, particularly in endemic regions, presents a promising path to reducing the incidence of leprosy and improving integration social status of
patients. This systematic literature review (SLR) examines the state of the art in research on the use of AI for leprosy care. From an initial 657 works from six scientific databases (ACM Digital Library, IEEE Xplore, PubMed, Scopus, Science Direct and Springer), only 30 relevant works were identified, after analysis of three independent reviewers. We have excluded works due duplication, couldn’t be retrieved and quality assessment. Results show that current research is focused primarily on the identification of symptoms using image based classification using three main techniques, neural networks, convolutional neural networks, and support vector machines; a small number of studies focus on other thematic areas of leprosy care. A comprehensive systematic approach
to research on the application of AI to leprosy care can make a meaningful contribution to a leprosy-free world and help deliver on the promise of the Sustainable Development Goals (SDG
An identity in transition: A longitudinal study examining students’ mathematical identity across the transition from primary to post-primary school
Full text linkThe transition from primary to post-primary school is a crucial time in a student’s journey as a mathematical learner. While this transition has been widely studied, much of the existing research, particularly in the Irish context, has focused on tracking mathematical performance or gathering insights from teachers. Little is known about what happens to students’ mathematical identities during this transition. This qualitative longitudinal study sought to address this gap, foregrounding student voice and examining their mathematical identity journeys.
Using a sociocultural theoretical lens, the research conceptualises mathematical identity as fluid, socially constructed and narratively expressed. 17 students were followed over an 20-month period, with their experiences captured through semi-structured interviews and personal journey graphs. A dual analysis was employed; the Listening Guide was used to conduct a voice-centred analysis of three illustrative student narratives, while thematic analysis was applied to data from all
17 students.
The findings highlight the transition as a multifaceted process involving overlapping “mini-transitions” that often extended beyond First Year. It emerged as a period of significant identity negotiation marked by confidence, doubt and adjustment, as students navigated through a complex array of changes. A pattern of silent acceptance emerged, with many students quietly adjusting to distinct changes in their learning experiences without voicing discomfort. Key influences on their
identity included the role of the teacher, assessment practices, peer comparisons and learning continuity. The study contributes to growing research on mathematical identity by illuminating its relational nature. It shows how students reconfigure their sense of self as mathematics learners in response to social and institutional shifts. The research calls for responsive pedagogies and more holistic transition supports that recognise identity work as central to mathematics learning. It calls
for educational policies that foreground student voice and position mathematical identity alongside mathematical achievement in terms of importance
Generation of Semantically Consistent Text and Its Evaluation
Full text linkGenerating semantically consistent text remains a crucial challenge in Natural Language Processing (NLP), especially in input-controlled settings such as data-to-text generation. Despite advances in neural language generation, current models often produce outputs that contain semantic inconsistencies with the input, including omissions, hallucinations, and distortions. These issues stem from the black-box, non-deterministic nature of neural models, which limits their interpretability and controllability. This thesis investigates how semantic consistency, the property that all information conveyed by the output is entailed by the input, can be more reliably evaluated and ultimately improved in generated text. Focusing on data-to-text generation, the thesis proposes that advancing semantic controllability requires a combination of principled formalisation of semantic errors and robust evaluation methods. First, it synthesises prior work on error annotation to develop a task-agnostic consensus taxonomy for semantic errors, distinguishing omission, addition, and substitution errors. Second, it applies this taxonomy in human evaluation studies, including span-based semantic error annotation to better understand the nature and reliability of human judgements. Third, it assesses the capacity of automatic evaluation methods, including lexical, embedding-based, and prompt-based approaches, to reflect human ratings of semantic consistency. In this context, the thesis explores LLM-as-judge setups for scoring system outputs and uses LLMs to sanity-check repeated human evaluation experiments, investigating their generalisability and reliability.
Together, these contributions provide new tools insights, approaches and results for the fine-grained assessment of semantic consistency and lay the groundwork for more interpretable, reliable, and scalable evaluation frameworks in neural language generation
Mortality Modelling for Actuarial applications: Challenges and Implications for Pension Sustainability and Insurance Pricing
Full text linkMortality forecasting and longevity risk are critical concerns for insurers, employers offering occupational pension schemes, and governments providing state pensions and social welfare benefits. Moreover, as many countries, including Ireland, experience
an ageing population, increased longevity raises pension costs, potentially undermining the sustainability of funding models used by both Companies and the state. However, accurately estimating mortality rates and forecasting future trends remains a significant challenge. For instance, the SARS-CoV-2 pandemic had a profound
impact on mortality rates from 2020-2022 with potential long-term effects. While the death toll of a country often serves as the ultimate indicator of pandemic impact, the calculation of excess deaths (deaths beyond expected levels) is complex and subjective. Seasonal mortality variations also pose risks, with factors such as
climate change, medical advances, and improvements in living standards disrupting established patterns, adding uncertainty for insurers and policymakers. Accurate mortality estimation for those retiring due to ill health is another complex issue.Ill-health retirement rates increase with age and are higher for females. As pension
schemes raise retirement ages in line with increasing life expectancy and more females join the workforce, the mortality of ill-health retirees will become increasingly important. This is an especially important consideration for employers providing
occupational pensions and insurers providing insurance for these benefits and/or selling impaired life annuities. Modelling these mortality challenges is particularly difficult for smaller populations, where mortality rates can be highly volatile. In Ireland and Northern Ireland, current national life tables have shown inconsistent
patterns, lack regional coherence, and exhibit volatile mortality trends. By jointly modelling mortality across different populations, the accuracy and coherence of forecasts can be improved. Joint modelling, or multi-population modelling, enhances the robustness of final graduated rates by allowing populations to borrow strength
from each other. This thesis provides insights into these evolving mortality challenges, contributing to more informed decision-making and ultimately improving the long-term sustainability of pension schemes (both private and public) and increasing the offering of accurately priced pension products
Burst agitation rate promotes sustained semicontinuous cultivation of filamentous fungi in stirred tank reactors
Full text linkEdible filamentous fungi L. edodes (shiitake mushrooms) were cultivated in submerged fermentation in stirred tank bioreactors (STR) both in batch and semicontinuous cultivation in a corn steep liquor (CSL) medium. The adjustment of a combination of constant impeller agitation speed, a short duration of a high-speed agitation (burst), and the frequency of
bursts improved biomass (cell dry weight (CDW) titre from 1.75 to 4.95 g/L in a 96-h batch cultivation. These bioreactor process conditions were applied to a semicontinuous culture strategy to produce similar biomass density at a dilution rate of 0.02 h−1 for up to 10 days without washout over the duration of the fermentation. An increase in the dilution rate above 0.02 h−1 resulted in washout of L. edodes over time. Using a richer growth medium through the addition of malt extract, peptone, and molasses allowed L. edodes to grow to 4.7 g/L at a dilution rate of 0.025 h−1 without washout. The maximum biomass productivity (396 mg CDW/h) of the semicontinuous cultivation (D=0.02 h−1) was 1.9-fold higher than the batch cultivation 206 mg CDW/hour. Use of the richer growth medium at D=0.025 h−1 improved biomass productivity further
to 470 mg/h. Glucans, known bioactives, were present in the fungal biomass at a maximum of 14% of the cell dry weight (CDW) with b-glucans predominating over a-glucans
Unveiling the Miniband Structure of Graphene MoiréSuperlattices via Gate-Dependent Terahertz Photocurrent Spectroscopy
Full text linkMoiré superlattices formed at the interface between stacked 2D atomic crystals offer limitless opportunities to design materials with widely tunable properties and engineer intriguing quantum phases of matter. However, despite progress, precise probing of the electronic states and
tantalizingly complex band textures of these systems remain
challenging. Here, we present gate-dependent terahertz photocurrent spectroscopy as a robust technique to detect, explore, and quantify intricate electronic properties in graphene moiré superlattices. Specifically, using terahertz light at different frequencies, we demonstrate distinct photocurrent regimes, evidencing the presence of avoided band crossings and tiny (∼1 to 20 meV) inversion-breaking global and local energy gaps in the miniband structure of minimally twisted graphene and
hexagonal boron nitride heterostructures, key information that is inaccessible by conventional electrical or optical techniques.
In the off-resonance regime, when the radiation energy is smaller than the gap values, enhanced zero-bias responsivities arise
in the system due to the lower Fermi velocities and specific valley degeneracies of the charge carriers subjected to moiré superlattice potentials. In stark contrast, the above-gap excitations give rise to bulk photocurrents intriguing optoelectronic responses related to the geometric Berry phase of the constituting electronic minibands. Besides their fundamental importance, these results place moirésuperlattices as promising material platforms for advanced, sensitive, and low-noise
terahertz detection applications
Scalable Lignin Monomer Production Via Machine Learning-Guided Reductive Catalytic Fractionation of Lignocellulose
Full text linkEfficient valorization of lignocellulosic biomass into high-value lignin
monomers is a cornerstone of sustainable biorefineries, yet the complexity of optimizing reductive catalytic fractionation limits industrial scalability. This study presents a machine learning (ML)-driven framework that harnesses 3,451 experimental data points from 54 peer-reviewed studies to model and optimize lignin monomer production. Among four advanced ML models developed, eXtreme Gradient Boosting Regression is found to achieve the highest predictive accuracy (R = 0.80–0.86) with low prediction errors (root mean square error: 3.99–8.31; mean absolute error: 2.85–6.90) for monomer production. Feature importance analysis reveals that operational parameters account for the largest influence (40–57%), followed by substrate content (25–43%) and catalyst-solvent properties (14–21%). The error between experimental and ML-predicted total monomer yields ranges from 2% to 2.6%, demonstrating robust performance of the model. Scaling this approach has the potential to process 140 million tons of aspen biomass annually, can
reduce CO2 emissions by 20.6 million tons, and yield $4,729 million in
socioeconomic savings. This ML-enhanced strategy offers a scalable and
environmentally viable pathway for data-driven lignocellulose valorization,
advancing the development of low-carbon, economically competitive
biorefineries
DCU MemoriEase at the NTCIR-18 Lifelog 6 Task
Full text linkWe present the participation of the MemoriEase lifelog retrieval system in the NTCIR-18 Lifelog 6 Task. This current MemoriEase system is an automatic and enhanced version of the MemoriEase system at the Lifelog Search Challenge 2024 (LSC). We report our methods for the two core sub-tasks in the NTCIR-18 Lifelog 6 task, Lifelog Semantic Access (LSAT) and Lifelog Question Answer (LQAT). We enhance the main architecture of the MemoriEase system utilizing the BLIP2 and CLIP embedding models to extract visual embedding and perform a comparison between the two models. In addition, we also use pseudo-relevance feedback for ad-hoc queries. For the LQAT sub-task, we use our retrieval model as the retriever and GPT-4o as a reader to generate answers to questions. Results of the LSAT sub-task show that our system found 369 images in 1,995 relevant images. The performance on known-item search queries is higher than on Ad-hoc queries, with 28.22% R@ 5 compared to 5.98% R@ 5, respectively. In the LQAT sub-task, the LLM model generates 8 correct answers in 24 questions. Although the performance is not high, it shows the advantages and drawbacks of the MemoriEase retrieval system and the QA model
Techno-economic analysis of green hydrogen storage in salt caverns: Evaluating cycling effects and cavern scaling on the levelized cost of hydrogen storage in Ireland’s power-to-X landscape
Full text linkThis paper examines the techno-economic feasibility of utilising salt caverns for large-scale hydrogen storage in Ireland, leveraging wind energy and proton exchange membrane (PEM) electrolysers. The analysis focuses on optimising the integration of wind power with hydrogen production and storage, addressing key challenges such
as energy curtailment, grid transmission constraints, and renewable energy intermittency. Findings highlight significant economic considerations, with a single hydrogen storage cavern requiring an initial investment of approximately €240 million, where geological site preparation and compressor systems constitute the largest cost
components. Annual operational expenses (OPEX) are estimated at €4.6 million, largely due to compressor energy consumption and cooling requirements. The study emphasizes the critical impact of electrolyser scale on economic viability. Small-scale systems, such as a 20 MW PEM electrolyser, are economically unfeasible, with a levelised cost of hydrogen (LCOH) of around €10/kg and filling times extending up to 2.5 years. However, scaling up to a 200 MW PEM electrolyser dramatically improves cost efficiency, lowering the LCOH to approximately
€0.83/kg and reducing filling times to just 90 days. This research provides a comprehensive framework for hydrogen storage development, offering key insights for policymakers and industry stakeholders to drive the
renewable energy transition and enhance energy security through cost-effective and sustainable storage solutions
Acute L-Carnitine Supplementation Does Not Improve CrossFit® Performance: A Randomized, Double-Blind, Placebo-Controlled Crossover Study
Full text linkBackground: L-carnitine supplementation is thought to enhance exercise performance, particularly in moderate and high-intensity activities, but evidence supporting this is mixed. This study aimed to assess whether acute L-carnitine tartrate supplementation could improve CrossFit® performance, specifically during the “Cindy” workout, a high-intensity
exercise protocol. Methods: In a randomized, double-blind, placebo-controlled crossover design, 20 trained male recreational CrossFit® athletes completed the “Cindy” workout within a 20 min period after ingesting either 3 g of L-carnitine tartrate or a placebo 90 min
before exercise. Performance was measured by total repetitions completed. Secondary outcomes included ratings of perceived exertion (RPE), gastrointestinal issues, and blood pressure (BP) measurements. Results: The results showed that L-carnitine supplementation did not significantly affect the number of repetitions performed (202.4 ± 69.9 vs.
204.5 ± 78.8, p = 0.810) compared to the placebo. There were also no significant differences in RPE (6.3 ± 1.5 vs. 6.9 ± 1.4, p = 0.180) or BP changes between groups. However, 10% of participants reported difficulty sleeping after L-carnitine supplementation. Conclusions: The findings suggest that 3 g of L-carnitine tartrate does not enhance CrossFit® performance in recreational athletes. Further research is needed to clarify its potential benefits, especially with larger samples and consideration of factors like sex and carbohydrate co-ingestion