16119 research outputs found
Sort by
CIRCLE: A framework for evaluating AI from a real-world lens
Most AI evaluations rely on static benchmarks that measure
model outputs in isolation, offering little evidence about how systems
behave once embedded in real-world workflows. As a result, decision-
makers lack systematic evidence about downstream effects, operational
risks, and long-term impacts that matter for deployment, governance, and procurement. We introduce CIRCLE, a six-stage lifecycle-based framework that links stakeholder concerns to context-sensitive evaluation methods, longitudinal measurement, and ongoing monitoring of deployed AI systems. The framework integrates evaluation methods such as A/B testing, field testing, red teaming, and longitudinal studies into a coordinated evaluation pipeline rather than treating them as isolated activities.
Together, these methods support more contextualized, iterative, and
decision-relevant assessments of AI systems. By aligning constructs, methods, and metrics with real deployment contexts, CIRCLE supports more actionable, iterative, and governance-relevant evaluation of AI systems and their secondary and tertiary effects
Contrast enhancement potential of food-based imaging agents in cholangiopancreatography and digestive tract investigations using MRI: A systematic literature review and meta-analysis
Objective
To evaluate the contrast enhancement potential of food-based substances (FBS) as clinical imaging agents in MR cholangiopancreatography (MRCP) and digestive tract investigations.
Methods
A systematic literature search was conducted using a pre-defined strategy and inclusion criteria to identify relevant articles. Two review authors independently screened and selected papers for inclusion, conducted risk of bias assessment using ROBINS-I (Risk of Bias In Non-randomised Studies of Interventions) tool and assessed certainty of evidence using the Cochrane’s GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach. Areas of disagreement were resolved through consensus and the involvement of a third reviewer. The data obtained were meta-analysed using both random and fixed effects models with inverse variance methods, depending on the included studies’ methodological heterogeneity, to estimate the pooled mean difference and odds ratio.
Results
16 studies (1360 participants) were included in this review. Eight studies each examined the effectiveness of FBS for contrast enhancement in MRCP and MRE, respectively. The pooled mean differences between MRCP with pineapple juice and non-contrast MRCP were 1.04 (95 % CI: 0.23, 1.84; I2 = 96 %, p < 0.01) for bile duct visibility and 0.95 (95 % CI: 0.04, 1.85; I2 = 95 %, p < 0.01) for pancreatic duct visibility. Compared to non-contrast MRCP, MRCP with pineapple juice showed improved pancreaticobiliary duct visibility (Combined odds ratio, 5.01; 95 % CI: 2.34, 10.64; p < 0.0001). However, when compared with synthetic contrast agents, there is a lower likelihood of obtaining excellent quality MRE images with food-based contrast agents (FBCAs) (odds ratio, 0.36; 95 % CI: 0.18, 0.73, 0.73; p = 0.004). A generally high safety and acceptance profile was reported across the included studies for FBCAs. The certainty of the evidence obtained was considered moderate across all outcomes.
Conclusion
Oral administration of pineapple juice improves the diagnostic quality of MRCP investigations, while the contrast enhancement potential of FBS is limited in MRE studies. The high heterogeneity of results and moderate certainty of the evidence, however, require a cautious admission of results. Therefore, further research is necessary to gain a comprehensive understanding of the reliability of FBS in these MRI investigations, as well as in other imaging modalities that are yet to be explored
The pressure record underfoot: Using ground penetrating radar to obtain pressure patterns of buried footprints
Fossilised trackways are a key source of evidence for reconstructing how ancient humans and animals moved across the landscape. The morphology of tracks is of particular importance because it allows inferences to be made about plantar pressure and therefore mechanics of motion of the track maker. However, the typical methods of excavation and photogrammetry used to document these tracks can be time consuming and may not be ideal at sensitive sites. This paper explores the use of ground penetrating radar to record the morphology of human tracks buried in soft sediment at White Sands National Park, New Mexico. The results demonstrate that a record of plantar pressure is preserved in the radar data and that this record might be a more direct measure of pressure than the typical proxy of footprint depth. This suggests that ground penetrating radar is a strong choice of method in ichnological studies
Effects of HPT on microstructure and mechanical properties of extruded secondary Al 6061 alloy
The influence of high pressure-torsion (HPT) on the microstructure and mechanical properties of extruded secondary aluminum 6061 alloy was investigated. Although previous HPT-processed materials have mainly been derived from primary wrought alloys, the novelty of the current study was derived from integrating HPT into the processing of secondary recycled wrought alloys. Mixed wrought aluminum scrap was hand-sorted to obtain a secondary aluminum with composition similar to Al 6061 wrought alloy. The resulting samples were extruded using a customised general-purpose pneumatic press. The extruded rods have a diameter of 14 mm. Discs with 10 mm diameter and 1.2 mm thickness were machined from the extruded rods and were further ground and polished to reduce disc thickness to 0.8 mm. The discs were then processed by HPT at 0.25, 0.5, 1, 2, 5 and 10 turns with 6.0 GPa constant pressure. Optical microscopy, SEM and EDS analysis, microhardness and tensile tests were conducted on the as-cast, as-extruded and HPT-processed disc samples. The results showed that HPT processing refined the grain size and increased microhardness from 50 to 110 HV. Tensile
test results demonstrated that HPT samples processed to 10 turns had UTS value of 381 MPa while non-HPT processed samples (as-cast and as-extruded) had UTS values of 84 and 214 MPa. These property improvements for secondary Al 6061 alloy opens up potential applications such as in micro-electro-mechanical systems
Between two worlds: Disjointed trajectories in the local and global circulation of Italian made-for-television horror series
The article addresses the challenges faced by filone filmmakers to adapt to televisual language and content in the 1980s. Made-for-television horror films exemplify this failed adaptation process in a decade characterized by the rise of commercial television and the consolidation of the dominant market position of Silvio Berlusconi’s broadcasting empire. They are also an example of the impact of distribution patterns on discursive practices surrounding the international circulation of Italian horror movies. Challenging ongoing scholarly debates in anglophone contexts, the article aims to shed light on the impact of transnational distribution practices peculiar to production companies like Dania Film and Medusa Distribuzione on the marketization of made-for-television horror series in anglophone countries and beyond. The combination of old and new industrial traditions may not have favoured the series’ success in Italy on the verge of the 1990s, but it ensured their circulation around the globe long after
Democracy, transparency, economic development, religion, and women on boards of national Olympic committees: Evidence from 89 countries
The authors investigate the connection between a country’s level of democracy, transparency, economic development, religion, and the representation of women on the national Olympic committees’ (NOC) boards. Using data from 89 countries, they analysed how democracy, governance quality, gross national income per capita, and religious composition are associated with women’s representation on NOC boards. Results reveal that countries with higher levels of democracy, greater transparency scores, higher economic development, a large proportion of Protestants, and a higher proportion of Jewish adherents were related to a greater proportion of women on NOC boards. The findings highlight the importance of democratic institutions, governance quality, and economic prosperity in creating conditions that support women’s advancement into decision-making roles, while religious context further influences cultural and normative factors that influence gender equality. Further research should explore the interplay among types of democracy, religious indices, and geographical characteristics to better understand the mechanism driving gender-balanced leadership. Overall, promoting democratic principles, participatory and inclusive governance, transparency, economic opportunity, and religious freedom can facilitate inclusive societal norms and enhance gender diversity, equity, and inclusion leadership roles
Mind the gap: gender equality and economic empowerment in the Commonwealth
This article examines gender inequality in economic participation across the Commonwealth countries from 2008 to 2023. Using secondary data, differences in factors such as wages, leadership, enterprise ownership, digital skills and legal protection are explored. Findings show that while legal reforms have been initiated in some countries, progress is uneven, particularly in low- and middle-income countries. Results indicate that a combination of legal frameworks, training and digital access supports gender equality. From the policy perspective, coordinated action to achieve Sustainable Development Goals (SDGs) 5 and 8 is suggested, as this will support female empowerment and equality across the Commonwealth countries
Wireless Vision-Centered Semantic Communication for smart city environment: Pretrained network and quantization
This paper introduces a Vision-Centered Semantic
Communication (VCSC) system tailored for efficient image transmission in smart city environments, where bandwidth is limited
and channels are subject to severe noise. Unlike conventional textcentered or classical compression approaches, VCSC leverages
a pretrained latent encoder–decoder network to extract compact, semantically rich representations directly from images. An
innovative attention-based quantization strategy is employed to
selectively allocate higher precision to critical regions, thereby
reducing the overall bit rate while preserving essential semantic
details. The quantized latent codes are robustly transmitted over
wireless channels modeled with additive white Gaussian noise and
Rayleigh fading. An end-to-end training framework minimizes
both reconstruction and perceptual losses, ensuring high-fidelity
image recovery even under adverse conditions. Extensive simulations demonstrate that VCSC outperforms traditional methods in
preserving fine-grained details and semantic integrity, offering a promising solution for real-time surveillance, transportation, and
infrastructure monitoring in smart cities
Factors influencing the adoption of management accounting practices in Cameroon.
Purpose: The aim of this paper is to identify factors that influence the adoption of these MAP by organisations in Cameroon (a former German, British and French colony) and how MAP can help organisations to be resilient to the internal and external environment.
Methodology/Approach: The research method used is case study and data was collected using 20 in-depth interviews with managers and accountants of private sector organisations in Cameroon, and document review.
Findings: There is considerable disparity in the adoption of MAP by organisations operating in Cameroon due to colonial heritage. Organisations in the English-speaking regions do consider professional accountants in contrast to those in the French Cameroon. Due to external factors such as the Anglophone crisis, government intervention through price regulation, and the COVID-19 pandemic, organisations have adopted innovative MAP to help them stay resilient.
Conclusion: MAP play a key role in decision making organisations. Unfortunately, in the case of Cameroon differences in the use of accounting professionals means there is no consistency in decision making. Government should put in place legislations and provide support for organisations to use accounting professionals to benefit organisations.
Practical implications: Management Accounting is one of the techniques used by organisations to govern their organisations and make good decisions. Therefore, understanding how these practices play in different socio-economic, cultural and political settings is of great importance. This is exactly what this research achieves.
Limitation: This research is limited to a few organisations in the two regions of the country. This means their perception and use of MAP might be different for other organisations facing different situations
Epidemic spread with asymptomatic infectious period in contact adaptive networks
We model an epidemic spread process involving nodes that (a) experience non-trivial asymptomatic infectious periods and (b) adapt by avoiding contacts with symptomatic infectious nodes. These modeling choices reflect ample evidence that infectious individuals are often mistakenly perceived as safe contacts due to lack of symptoms and that individuals adapt to an epidemic by avoiding contacts deemed to be of risk. We capture these choices in the SIa Is S (Susceptible-Asymptomatic InfectedSymptomatic Infected-Susceptible) model, where we explicitly distinguish between asymptomatic and symptomatic infectious
individuals. We adopt an individual-based mean-field epidemic modeling approach and formulate the system of differential equations via continuous-time Markov chain analysis. We first consider non-adaptive homogeneous and heterogeneous mixing scenarios over arbitrary static contact networks. We derive the expression for the basic reproduction number, R0, and establish that under otherwise similar conditions the individual infection probabilities at the metastable state of SIa Is S dominate those in the conventional SIS model. Then, we focus on a contact-adaptive setting, where nodes avoid interactions with
known infectious neighbors or reconnect with neighbors who have recovered, to study how the time-varying contact network, asymptomatic infections and their combinations affect the epidemic spread dynamics. Overall, asymptomatic infections restrict nodes’ capacity to adapt (link breaking), resulting in higher link density and consequently higher epidemic prevalence. Besides, in their presence, the retarding effect of the link-breaking mechanism on epidemic prevalence is considerably mitigated. We numerically analyze how the effective link-breaking rate and the size of the asymptomatically infected population affect the
link density and, ultimately, the epidemic prevalence. Furthermore, the epidemic threshold appears to scale inversely with the population of asymptomatic nodes, namely the epidemic starts to spread at lower infection rates when the number of asymptomatic infections is higher