Capsule endoscopy (CE) enables noninvasive visualization of the gastrointestinal tract but generates tens of thousands of images per examination, creating substantial interpretive burden, reader fatigue, and uncertainty in lesion localization. Artificial intelligence (AI), particularly deep learning-based image analysis, has markedly improved lesion detection and reduced reading time; however, image-only models remain limited in anatomical localization and spatiotemporal context. This narrative review summarizes current evidence on multimodal AI in CE, focusing on systems integrating capsule video with sensor-derived data such as inertial, magnetic, and physiological inputs. A structured search was conducted. Image-only convolutional neural network models consistently demonstrated high sensitivity (> 95%) for detecting gastrointestinal bleeding, ulcers, vascular lesions, polyps, and neoplasia, while reducing mean reading time from approximately 40-60 minutes to under 5 minutes. Multimodal architectures-including convolutional neural network-long short-term memory hybrids and sensor-fusion frameworks-further enabled accurate organ classification, transit time estimation, and improved lesion localization, particularly under suboptimal image quality. Overall, image-only AI excels in visual lesion detection, whereas multimodal AI enhances spatial awareness and diagnostic context. Key challenges include sensor heterogeneity, limited prospective validation, and regulatory considerations. With standardization and large-scale validation, multimodal AI may enable more context-aware and clinically interpretable CE.

ABSTRACT The caste system in India is based on occupational inheritance and a restricted social hierarchy in which the birth of a particular caste determines its position in society. In the modern era, individuals’ identity is defined by their caste. Among the lowest castes, severe illiteracy and a lack of information about support services leave many people suffering in an increasingly complicated technological environment. They frequently feel disoriented and struggle to understand the intricacies of the contemporary technological world. The determined efforts of Dalit activists stand as beacons of light in these difficult conditions. Dalit activists are pivotal in addressing these challenges, dedicated to confronting systemic inequities and championing marginalized voices for positive change. They provide support to the people who are marginalized, making sure that their issues are recognized and taken up in a larger public forum. This research examines how Dalit women activists use the social media platform X to advocate for Dalit rights, engage in anti-caste conversations, and build solidarity against long-standing caste prejudice. This study employs critical caste studies as a lens to study the phenomenon of Dalit activism on social media. Netnography is the technique used, with participants being observed and concentrating on online fieldwork.

Many liquid metal plasma-facing component concepts for fusion reactors require a capillary porous material to wick the liquid metal to the surface of the component. In this work, samples of 3D-printed molybdenum lattices designed to have pore sizes ranging from 100 to 1000 µm were fabricated by two different suppliers. The dimensions of the lattice structures were characterized using an optical microscope, revealing that the lattice walls were almost systematically thicker than they were designed to be and in turn the pores are all smaller than they were designed to be. In many cases the pores were completely clogged. Several spectroscopic techniques were used to characterize the chemical composition of the materials, collectively indicating the presence of a thick layer of oxygen-rich material extending more than 1 µm below the surface. Compression testing revealed significant asymmetry in the mechanical behavior depending on which direction the force was applied relative to the print direction. Lastly, lithium wetting tests of several lattice samples were performed, and a sample with ~250-µm pores was found to soak up lithium quickly after initial wetting of the surface occurred. The implications of these results for future liquid metal plasma-facing component design are discussed.

This paper proposes a hybrid speed–current control framework for high-performance induction motor drives under field-oriented control. The outer speed loop combines Backstepping, Radial Basis Function neural approximation, and high-order sliding mode compensation, where Backstepping provides a Lyapunov-based nonlinear structure, the RBF network estimates lumped uncertainties online, and the HOSM term ensures finite-time disturbance rejection and fast convergence. An enhanced Harris Hawks Optimization with chaotic Lévy flights is employed to optimally tune the controller parameters. In the inner loop, a Higher-Order Non-Singular Terminal Sliding Mode controller regulates stator currents, guaranteeing finite-time tracking, reduced chattering, and robustness against parameter variations and inverter nonlinearities. Lyapunov analysis confirms uniform ultimate boundedness of the speed loop and finite-time convergence of the current loop. The effectiveness of the proposed strategy is verified through MATLAB/Simulink simulations and real-time implementation on the OPAL-RT OP5707XG platform, demonstrating superior dynamic performance and disturbance rejection compared with PI- and conventional sliding-mode-based schemes.

ABSTRACT This study examines pre-service geography teachers’ perceptions of food systems, a newly introduced topic in South Korea’s 2022 revised geography curriculum. Using Q methodology, four distinct perception types were identified: Technological Innovation Pragmatists, Ethical Autonomy Practitioners, Integrative Education & Sustainability Advocates, and Contextual Balance Selectors. To enhance the strengths and address the limitations of each type, this paper proposes targeted pedagogical strategies: critical case studies to foster multi-dimensional understanding (Type 1); structural connection activities to link ethical consumption with political-economic contexts (Type 2); methodology-focused training to apply concept mapping and problem-based learning (Type 3); and decision-oriented discussions to enhance rational judgment (Type 4). These findings offer a diagnostic framework and practical prescriptions for teacher education programs, contributing to the cultivation of teachers capable of fostering food citizenship and sustainability.

Purpose: This study explored how Lake Victoria Basin Water Board’s (LVBWB) social media practices influence public awareness, perceptions and conservation-related intentions among stakeholders in Mwanza City, Tanzania. The study interrogated platform usage, content characteristics, engagement dynamics and institutional practices to understand why digital visibility often fails to translate into sustained conservation behavior. Methodology: Purposive qualitative case study design was employed. Data sources included documentary review of LVBWB social media outputs and institutional materials (March–September 2025), non-participant observation at two outreach events and Key Informant Interviews from 14 purposively sampled key-informant interviewees with LVBWB staff, community leaders, fishing community representatives, youth group leaders and Non-Governmental Organisation environmental actors. Then data were grouped into themes and subjected to content analysis. Anonymized illustrative excerpts are included. Findings: Five interrelated themes emerged: (1) episodic and event-driven posting that produces transient salience rather than cumulative agenda-setting; (2) language and tone barriers, where formal or technical citations limit accessibility; (3) visibility without dialogue, where “likes” and impressions do not equate to participation or ownership; (4) local intermediaries as trust multipliers, with community leaders and youth champions essential for mobilization; and (5) institutional constraints (approval bottlenecks, limited communication capacity, and absence of formal feedback protocols) that inhibit two-way engagement. The study proposes a pragmatic hybrid communication model combining sustained social media practice, Kiswahili-first messaging, local co-creation and institutionalized feedback to bridge awareness–action gap. Unique contribution: This paper provides contextually grounded, actionable guidance for LVBWB and similar institutions in East Africa by integrating TAM-informed adoption insights with Public Relations and social marketing principles, while retaining strict fidelity to qualitative evidence.

ABSTRACT Food safety issues have attracted increasing public attention and can affect not only domestic food consumption but also international trade in agricultural products. This paper examines the impact of the 2011 lean meat powder scandal in China on pork imports. Exploiting the exogenous exposure of the scandal, we employ a difference-in-differences approach comparing China’s pork imports from economies that allow lean meat powder with those that prohibit it, thereby constructing a quasi-natural experiment. To assess robustness and address substitution concerns, we also apply the synthetic control method to construct counterfactual import trajectories. The results show that, after the scandal, pork import quantities from the treatment group declined by 24.9%, while import values fell by 35.0%. The negative impact is more pronounced and persistent for offal products associated with higher residual risks. Supplementary evidence based on information search intensity suggests that heightened public attention is associated with stronger import reductions, consistent with a trust-related transmission mechanism. From a policy perspective, the findings indicate that food safety scandals can generate lasting disruptions in international agricultural trade, underscoring the importance for exporting economies of maintaining credible food safety standards and monitoring conditions in importing economies to mitigate trade risks from food safety shocks.

ABSTRACT Taphonomic studies conducted thus far at Orce sites (Guadix-Baza Basin, Granada, Spain) have not provided definitive conclusions regarding early human subsistence strategies. Recent research has neither confirmed nor ruled out early human access to the carcasses of animals recovered at these sites. To expand the available evidence, this study proposes a comparative analysis of mortality patterns at these Early Pleistocene locations, alongside published data on modern and fossil accumulations. Mortality patterns have been widely used to analyse aspects of human and animal populations, including subsistence strategies, seasonal variations, and ethology. These patterns consider the proportional representation of different age cohorts to infer the accumulation processes and causes of mortality in fossil communities. In that sense, mortality profiles from Venta Micena-3 and Venta Micena-4 align with their respective interpretations of Pachycrocuta brevirostris den and mortality scenario where carnivores preyed. Also, Barranco León and Fuente Nueva-3 show an admixture of different situations, though their respective sample sizes are not decisive. In conclusion, while the results of this study are coherent with previous interpretations, they prove to need further analyses to increase the sample and open potential scenarios relevant to the Early Pleistocene sites of Orce. Zoobank ID: urn:lsid:zoobank.org:pub:5F55E469-38CC-46F4-8EFF-6A037E4D8B64