Browse Articles

Reforming disease prognosis and treatment prediction for palliative care with hybrid metaheuristic deep neural architectures in IoT healthcare ecosystems

Evaluating the clinical utility of multimodal large language models in rare maculopathy

Abstract This study aimed to assess how multimodal large language models (MLLM) diagnose and differentiate Pentosan Polysulfate (PPS) Maculopathy from other phenotypic mimics. A retrospective review of clinical records and multimodal retinal imaging was conducted with patients from the Shiley Eye Institute and Casey Eye Institute. Four MLLMs (ChatGPT-4o, Claude 3.5 Sonnet, Google Gemini 1.5 Pro, Perplexity Llama 3.1 Sonar/Default) along with human retinal specialists answered prompts based on retinal imaging and demographic data. Performance was evaluated using accuracy, sensitivity and specificity estimates. The study included 126 eyes from 63 patients, with 36 eyes with PPS maculopathy, 50 eyes with Stargardt disease, and 40 eyes with PRPH2 -associated multifocal pattern dystrophy. MLLMs showed improved accuracy and sensitivity when answer choices were restricted, with ChatGPT consistently performing best when all imaging modalities were prompted together. The inclusion of demographic data further enhanced performance in prompts with limited answer choices. Human retinal specialist evaluations aligned with MLLM performance trends and also improved with demographic data. While MLLMs show diagnostic potential, further refinement is needed before clinical implementation. These findings highlight the importance of prompt design and demographic data to optimize MLLM performance with retinal imaging modalities.

A high utility itemsets mining algorithm based on co-evolution

A bothy among the stars

Modified Mann-Kendall with higher-order statistics for trend analysis

Headache prevalence and impact among school-aged children in a Japanese town: the AMI-GRAINES study

Abstract Headaches are common in youth and can impact school and daily life. Data on headaches and care-seeking in Japanese schoolchildren are scarce. We investigated prevalence, impact, and consultation patterns. A school-based online survey was conducted between February and March 2025 among 3,766 students in seven elementary and three junior high schools in Ami-town, Ibaraki-prefecture, Japan. Two questionnaire versions were distributed: one for lower-grade students (grades 1–3) and one for upper-grade students (grades 4–9). The survey included items on headache frequency, triggers, associated symptoms, medication use, and consultation experiences. The prevalence of headaches was 29.1% in lower-grade and 37.6% in upper-grade students. The most common age of onset was 10 years. Students with Headaches that meet most of the diagnostic criteria for episodic migraine experienced more headache days and more school absences compared to students with other headache characteristics. Although nearly half of the students had taken medication, fewer than one-third had consulted a physician. Despite experiencing headaches, only a small number of children sought medical care, which may raise concerns about the potential overuse of over-the-counter medications. Therefore, collaboration between educational and medical institutions is essential to improve early recognition and intervention.

Propolis impairs cellular proliferation by promoting oxidative damage and disrupting macromolecular composition in a yeast model

Research on the thermal response characteristics of building nano bead reflective insulation materials based on multi-physics field coupling

Abstract In this study, to advance the deployment of nano‑microbead reflective insulation in green, low‑carbon buildings, we developed a three‑dimensional mesoscopic model with randomly distributed nano‑beads in COMSOL Multiphysics. Using a coupled radiation–conduction approach, we systematically compared the outer–inner surface temperature difference under radiative‑only, conductive‑only, and fully coupled conditions as a function of solar irradiance. The coupled scenario exhibited the smallest temperature-rise slope, which supports improved insulation performance under the investigated conditions. Sensitivity analysis showed that a bead diameter of 100 μm and an 88% porosity minimize the equivalent thermal conductivity while ensuring adequate mechanical strength. Under a diurnal cycle of solar loading and natural convection, dynamic experiments revealed thermal relaxation times of approximately 2 h (outer surface) and 5 h (inner surface), and quantified a 3.0 °C reduction in outer‑surface temperature per 1 W/(m²·K) increase in the convection coefficient. These behaviors reflect a pseudo‑linear thermal response arising from the linearization of radiative heat transfer and homogeneity assumptions. Finally, we propose optimization strategies incorporating dynamic radiation, enhanced convective dissipation, and temperature‑dependent material properties to guide the design of high‑performance nano‑bead reflective insulation for building envelopes and photovoltaic applications.

Experimental and theoretical insight into the complexation of tetra secondary butyl diglycolamide (TSBDGA) with trivalent f cations into ionic liquid

An in vitro accuracy study on scan body-assisted surface based registration for conventional and zygomatic dental implants

A linear-attention based network for estimating continuous upper limb movement from surface electromyography

Order in which cancer-driving mutations occur affects the chance of tumour development

AI-derived five-gene signature predicts risk in multiple myeloma under bortezomib-based therapy

Reconfiguration of DNA methylation at growth resumption after winter dormancy in Kiwifruit

Bifunctional LYTAC Mediates Hepatocytes-Hepatic Stellate Cells Crosstalk by Regulating 5-HT _2A Receptor Degradation and Antagonism to Synergistically Ameliorate Hepatic Fibrosis

Hybrid island-and-sea approach for corrosion protection of Si photocathode in neutral-pH water splitting

Abstract Photoelectrochemical (PEC) water splitting is a promising way for converting solar energy into green hydrogen, yet the long-term stability of the studied photoelectrodes remains a main challenge. Therefore, corrosion protection of photocathode and photoanode materials has attracted attention. Here, we demonstrate a simple yet effective “island-and-sea” strategy to enhance the stability of silicon-based photocathodes in neutral media water splitting. Platinum nanoparticles (“islands”) deposited on Si facilitate efficient charge transfer, whereas the remaining surface is passivated with a hydrophobic 1-octadecyl (OD) self-assembled monolayer (“sea”) that acts as a corrosion-resistant barrier. This organic-protective layer allows stable PEC operation without altering the semiconductor’s band structure and complex fabrication steps. The Si/Pt + OD structure maintains a stable photocurrent over 6 h, compared to a 50% decline after 3 h of operation for the unprotected Si/Pt photocathode. Since neutral electrolytes like seawater are readily available and inexpensive natural resources on the planet, the “island-and-sea” approach applicable for stable operation at neutral pH becomes crucial for real-world applications.

Exploring CO2 solubility in 1-N-butyl-3-methylimidazolium hexafluorophosphate ionic liquid using neural network models

Fear of disease progression and related factors among chronic disease patients attending South Wollo zone government hospitals

Analyzing the impact of a discounted parameter on the reduction of collision time in Brownian particle trajectories

Deep learning approach for crop-weed segmentation in peanut cultivation using PSPEdgeWeedNet

Abstract Weed management continues to be a significant challenge in modern agriculture, primarily due to the aggressive growth patterns of weeds and their direct competition with crops for essential resources such as light, water, and nutrients. Although recent developments in precision agriculture have led to the emergence of automated weed detection systems aimed at reducing operational costs and decreasing reliance on chemical herbicides, achieving accurate crop–weed segmentation remains a persistent difficulty. This is largely attributed to high visual similarity between crops and weeds, coupled with variations in illumination and field conditions. To address these challenges, Convolutional Neural Networks (CNNs) have been increasingly adopted for their capability to perform end-to-end, pixel-level classification, particularly when leveraging multi-spectral imagery. In this context, PSPEdgeWeedNet is proposed, a novel edge-aware deep learning architecture tailored for precise semantic segmentation of crops and weeds within peanut cultivation fields. Distinct from the conventional Pyramid Scene Parsing Network (PSPNet) and its boundary-aware variant developed as a baseline in this research, PSPEdgeWeedNet introduces a dedicated edge detection branch. This branch is specifically engineered to enhance boundary localization and improve delineation between adjacent vegetation classes. In post-processing, Conditional Random Fields (CRFs) are used to slightly enhance the segmentation results around object boundaries. Additionally, all models were trained on a curated peanut field dataset using class-weighted loss functions to effectively address inherent class imbalance. Comprehensive experimental evaluations reveal that PSPEdgeWeedNet significantly outperforms existing state-of-the-art architectures including PSPNet, SegNet, UNet, DeepLabv3, Swin-Unet, and light weight transformer model based on ViT across multiple performance metrics such as Intersection over Union (IoU), precision, recall, and F1-score. These results highlight the critical role of incorporating edge-aware mechanisms within semantic segmentation frameworks, thereby enhancing the robustness and accuracy of automated weed detection systems in complex, real-world agricultural environments.