The brain clock, located in the suprachiasmatic nucleus (SCN) of the hypothalamus receives direct retinal input providing the entire body with an internal representation of external solar time. The pathways by which this small nucleus signals so broadly involve co-occurring nervous and diffusible output signals, though the latter are less understood. Portal pathways, such as the well-known pituitary portal pathway, provide a mechanism whereby signals of neural origin can reach local, specialized targets without suffering dilution in the systemic blood supply. The newly discovered pathways involve direct connections between each of the sensory circumventricular nuclei at their point of their attachment to the brain. These nuclei line the brain’s ventricles, and their leaky blood vessels and large perivascular spaces represent a route whereby secretions from the SCN can be relayed and then amplified, providing a pathway to achieve global coordination of circadian clock signaling. This review provides a narrative that incorporates our understanding of SCN neural and diffusible output signals, with particular emphasis on the contribution of brain fluid compartments and the fluids therein.

Title: Localized scleroderma and excimer laser: two case reports and a literature review

Case PresentationA 72-year-old woman with a history of uterine prolapse presented to the emergency department with a fever and malaise that persisted for several days. The patient had well-controlled diabetes mellitus. The patient’s temperature was 39.7℃, and her respiratory rate was 24/min; the other vital signs were unremarkable. On physical examination, no findings suggestive of an infectious source were noted. Blood test results showed a normal white blood cell count of 7,309/μL, but an elevated C-reactive protein level of 6.07 mg/dL. Additionally, her renal function was decreased, with a creatinine level of 1.47 mg/dL. Pyuria was confirmed to be present. Contrast-enhanced abdominal computed tomography revealed bilateral hydronephrosis and ureteral obstruction due to compression by uterine prolapse (Figure 1). A urine culture tested positive forEscherichia coli . Thus, obstructive urinary tract infection associated with uterine prolapse was diagnosed. The uterine prolapse was repaired by inserting a pessary, which was performed by gynecologist, and antimicrobial therapy was initiated. On day 2, ultrasound examination showed resolution of the hydronephrosis. On day 3, the fever resolved, and the patient’s creatinine level improved (0.97 mg/dL). Antimicrobial therapy was completed in 10 days. The patient was discharged on day 13.Uterine prolapse is typically asymptomatic and rarely requires emergency interventions.1 However, it can be complicated by obstructive urinary tract infections and severe renal dysfunction.2-3 Although its treatment requires removal of the obstruction, few reports on obstructive urinary tract infections due to uterine prolapse exist, and management strategies vary. Previous reports have shown that urologic intervention is not required to release the obstruction, and pessary insertion can be effective. 3 Once the uterine prolapse is corrected, antimicrobial therapy is sufficient for improvement. In our case, pessary insertion was also effective, as evidenced by the rapid amelioration of hydronephrosis. Emergency physicians should be aware that uterine prolapse is a risk for obstructive urinary tract infections or renal failure and that pessary insertions can be effective.

Diabetes, a pervasive chronic disease, characterized by insufficient insulin production or the body’s inefficiency in insulin utilization. With rising global spread and severe consequences like blindness, kidney failure, and stroke, timely detection is paramount. This paper introduces an innovative framework for diabetes detection using machine learning, concentrating on a benchmark dataset in the field, Pima Indians Diabetes Database. The dataset inherent challenges like class imbalance and missing values are dealt with utilizing Conditional Tabular Generative Adversarial network (CTGAN), and pre-processing methods. Furthermore, the study also employs an ensemble approach, combining four base models—Random Forest (RF), Logistic Regression (LR), Gaussian Naive Bayes (GNB), and K-Nearest-Neighbor (KNN)—trained on a balanced dataset and amalgamated through stacking with an Extreme Gradient Boosting (XGB) meta-classifier. The resulting ensemble model demonstrates superior performance, achieving 96% accuracy on the test set. In comparison, standalone models, exhibit lower accuracy at 85% on an average. This work highlights the effectiveness of ensemble techniques and data synthesis in improving diabetes prediction, and emphasizes the significance of early detection in mitigating the global impact of this life-threatening disease.

1. Farhad Ghorbani, Assistant professor of oral and maxillofacial surgery, Shiraz university of medical sciences. Shiraz, Iran.2. Jamaledin Motazedian, Postgraduate student of oral and maxillofacial surgery, Shiraz university of medical sciences. Shiraz, iran.3. Mohammad Saleh Khaghaninejad. Assistant professor of oral and maxillofacial surgery, Shiraz university of medical sciences. Shiraz, iran.4. Haleh Keshvari, Oral and maxillofacial pathologist. Rajaei hospital. Shiraz, Iran.5. Maryam Paknahad, Oral and dental disease research center, oral and maxillofacial radiology department, Shiraz, IranCorresponding Author: Dr. Maryam Paknahad Associate Professor of Oral and Maxillofacial Radiology Shiraz University of Medical Sciences Shiraz, Iran Email: Paknahadmaryam@yahoo.comCentral Odontogenic Fibroma: A Rare Benign Tumor with Potentially Life-Threatening Implications :A Case Report

A 32-year-old man with an uncommon cardiomyopathy

Endomyocardial fibrosis (EMF) is the most common restrictive cardiomyopathy worldwide. However, its etiology and pathogenesis remain unclear. Epidemiology data are limited but the highest prevalence has been reported in tropical region specifically sub-Saharan Africa (1). EMF is characterized by apical filling with fibrotic tissue in one or both ventricles, often associated with thrombus and calcification. Most often it extends to the posterior leaflet of the mitral valve, papillary muscle or chordae tendineae, causing atrioventricular valve dysfunction. Transthoracic echocardiography (TTE) is the first-line modality for assessment, but MRI has emerged as a more accurate tool for the tissue characterization of this disease(2).

Blood is significant biological evidence in forensic investigations, and before DNA analysis, it is frequently screened to verify its biological provenance. These presumptive tests use hydrogen peroxide (H 2O 2) as an oxidizing agent in conjunction with reagents such as Phenolphthalein (Php) and tetramethylbenzidine (TMB) to measure hemoglobin’s peroxidase activity. Although these compounds help detect the presence of blood, their genotoxic effects—mainly when H 2O 2 is present—have raised concerns about possible DNA destruction. This study uses the Comet test, a single-cell gel electrophoresis method that assesses DNA strand breakage without DNA extraction, to examine the genotoxic effects of TMB and Php on blood samples. Blood samples were subjected to different combinations of TMB, Php, and H 2O 2 to evaluate their impact on DNA integrity. The results showed that combining blood enhancement reagents and H 2O 2 caused oxidative DNA damage, forming 8-oxo-deoxyguanosine, a hallmark of DNA adducts. This study emphasizes that blood augmentation drugs should be carefully considered because they can degrade DNA, which could jeopardize forensic DNA analysis.

A document by Jordan Llerena-Velastegui. Click on the document to view its contents.

Time series classification is becoming more and more significant in many domains, yet current approaches frequently place computing economy below classification accuracy. In order to improve predictive performance while preserving computational efficiency, this study introduces Diverse Representative Time Series Features (DRTSF), a novel feature-based framework that integrates a wide range of transformations, including derivatives, Hilbert transform, discrete wavelet transform, fast Fourier transform, discrete cosine transform, and autocorrelation. DRTSF seeks to identify intricate patterns in time series data by merging several distinct representations into a single feature vector. Using all 142 datasets from the UCR collection, our analysis indicates that DRTSF outperforms top feature-based classifiers like FreshPRINCE. The results suggest that DRTSF could be especially useful for large-scale data analysis, providing a speed-accuracy trade-off that is similar to the cutting-edge Quant method. In contemporary data-driven applications, these findings underscore the possible benefits of combining several representations to handle the trade-off between computational cost and classification precision.

This paper introduces a novel Hamiltonian-inspired neural network approach to credit scoring, designed to address the challenges of class imbalance and out-of-time (OOT) prediction in financial risk management. Drawing from concepts in Hamiltonian mechanics, we develop a symplectic optimizer and a new loss function to capture the complex dynamics of credit risk evolution. Using the Freddie Mac Single-Family Loan-Level Dataset, we evaluate our model's performance against other machine learning approaches. Our method shows superior discriminative power in OOT scenarios, as measured by the Area Under the Curve (AUC), indicating better ranking ability and robustness to class imbalance. The Hamiltonian-inspired approach shows particular strength in maintaining consistent performance between in-sample and OOT test sets, suggesting improved generalization to future, unseen data. These findings suggest that physics-inspired techniques offer a promising direction for developing more robust and reliable credit scoring models, particularly in uncertain economic situations.

This paper introduces an innovative approach to analyzing and improving multi-hop reasoning in AI systems by drawing inspiration from Hamiltonian mechanics. We propose a novel framework that maps reasoning chains in embedding spaces to Hamiltonian systems, allowing us to leverage powerful analytical tools from classical physics. Our method defines a Hamiltonian function that balances the progression of reasoning (kinetic energy) against the relevance to the question at hand (potential energy). Using this framework, we analyze a large dataset of reasoning chains from a multihop question-answering task, revealing intriguing patterns that distinguish valid from invalid reasoning. We show that valid reasoning chains have lower Hamiltonian energy and move in ways that make the best trade-off between getting more information and answering the right question. Furthermore, we demonstrate the application of this framework to steer the creation of more efficient reasoning algorithms within AI systems. Our results not only provide new insights into the nature of valid reasoning but also open up exciting possibilities for physics-inspired approaches to understanding and improving artificial intelligence.

A knowledge graph is a multidirectional labelled graph used for the graphical representation of knowledge where each node represents an entity, and the edge connecting the two nodes represents a relationship. There has been a rise in the popularity of using knowledge graph in information retrieval, recommender system, dialogue system, and question-answering system. The question-answering system can be either over structured data or unstructured data. This article studies the existing techniques for questionanswering systems over knowledge graph. We collect 43 articles on the question-answering system over a knowledge graph. We give a brief introduction to knowledge graph and question-answering system. Further, we discuss challenges, datasets and evaluation metrics used to evaluate techniques. Finally, we conclude the article by discussing some open research aspects, highlighting factors of the low-resource language for question-answering systems over knowledge graph and remarks on existing systems.

Large language models have emergent capabilities that come unexpectedly at scale, but we need a theoretical framework to explain why and how they emerge. We prove that language models are actually non-ergodic systems while providing a mathematical framework based on Stuart Kauffman's theory of the adjacent possible (TAP) to explain capability emergence. Our resource-constrained TAP equation demonstrates how architectural, training, and contextual constraints interact to shape model capabilities through phase transitions in semantic space. We prove through experiments with three different language models that capacities emerge through discrete transitions guided by constraint interactions and pathdependent exploration. This framework provides a theoretical basis for understanding emergence in language models and guides the development of architectures that can guide capability emergence.

A regime shift is an abrupt, substantial, and persistent change in the state of a system. We show that a regime shift in the September Arctic sea-ice extent (SIE) occurred in 2007. Before 2007, September SIE was declining approximately linearly. In September 2007, SIE had its largest year-to-year drop (by a wide margin) in the entire 46-year satellite record (1979-2024). Since 2007, September SIE has been approximately constant, i.e., no long-term trend. The regime shift in 2007 was caused by significant export and melt of older and thicker sea ice over the previous 2 to 3 years, as documented in other studies. We test alternatives to the traditional linear model of declining September SIE, and discuss possible explanations for the lack of a trend since 2007.

A document by Xavier Pardell Peña. Click on the document to view its contents.

Real-time flood prediction is crucial for coastal urban cities prone to flooding; enabling communities, emergency services, and transportation authorities to prepare and take necessary precautions. Traditional approaches for flood prediction involve high-fidelity physics-based hydrodynamic models, which are computationally expensive, particularly in terms of time. To address these limitations, data-driven machine learning models have been developed for real-time predictions, focusing primarily on specific spatial dimensions such as street segments, floodplain areas, or geographic locations. In this study, we propose a generalized machine learning model based on Fourier Neural Operators (FNO), capable of operating across geographic locations not seen during training. For this research, we analyze eleven storm events in Norfolk, VA, spanning from 2017 to 2022, each lasting approximately four days with a 15-minute time interval. Our FNO model utilizes water depth maps generated by the TUFLOW hydrodynamic model, with a spatial resolution of 2.5m x 2.5m, along with rainfall data from seven observation sites maintained by the Hampton Roads Sanitation District. We employ inverse-distance weighted interpolation to account for geographic variations of rainfall across the study area. For model evaluation and generalization, we implement a k-fold cross-validation approach, randomly dividing the study area into five folds. Our findings show that training for 24 time steps (360 minutes) result in a model capable of accurate predictions in the next 6 time steps (90 minutes). Additionally, sequential experiments using k-fold with eleven storm event folds demonstrate that a look-back period of 360 minutes yields similarly low error rate predictions in the 15 minutes look-ahead interval. These findings underscore the efficiency and generalizability of our FNO-based model, demonstrating its capability to effectively handle predictions over extended periods for unseen geographic locations. Additionally, our machine learning model achieves an order of magnitude speed up compared to traditional physics-based hydrodynamic models.

Real-time, interactive 4D traffic scene generation enables rapid digital twinning of traffic scenarios, improving management and decision-making in intelligent transportation systems. However, current text-to-video models, such as Sora, struggle to maintain the temporal coherence of traffic elements and interact with dynamic environments and users when generating 4D scenes. This article introduces a novel cloud-edge-terminal collaborative framework that leverages Artificial Intelligence-Generated Content (AIGC) in vehicular networks to tackle these challenges, ensuring long-term coherence and improved interactivity. The framework presents a comprehensive architecture for real-time interactive 4D scene generation, encompassing data collection, management, model pre-training, fine-tuning, and inference. We examine key design requirements and challenges, demonstrating that our microservice-based framework enables the system to generate and update 4D traffic scenes in real time, effectively responding to traffic data and user inputs. To the best of our knowledge, this is the first successful implementation of real-time, interactive 4D traffic scene generation. Performance evaluations show the superiority of our framework, powered by microservice-based code fine-tuning, over traditional frameworks. Finally, we discuss future research directions to enhance AIGC-driven 4D traffic scene generation.

Urban ponds serve as important objects both for ecological research - spanning biodiversity conservation, landscape connectivity, ecosystem properties - and for studies on evolutionary dynamics due to anthropogenic stressors driving rapid adaptation in these ecosystems. They are strongly connected to human society and these interdisciplinary connections can be tackled within the ecosystem services - disservices framework. Garden ponds, a specific subset of urban ponds, have mostly been overlooked elements of urban landscapes, even though they might fulfil some of the same ecosystem roles as their generally larger counterparts and may even be more numerous in many areas worldwide. By synthesising knowledge on urban ponds, distinguishing garden ponds as a specific subtype, we highlight their ecosystem services (i.e., nature's contributions to people, NCPs) and disservices, and their connections to biodiversity conservation. We also identify potential ecosystem service multifunctionality and trade-offs that might be tackled by future research in the context of ecosystem services, climate change, and urban sustainability.

Level: Grade 1

We report MMS observations of typically unipolar Electrostatic Solitary Waves (ESWs) with amplitudes up to 100 mV/m in the Central Plasma Sheet (CPS) ($|B_x|<5$ nT) of the Earth’s magnetotail, outside the primary magnetic reconnection region during the intervals of Bursty Bulk Flows (BBFs). The generation of ESWs leads to a parallel electron temperature drop of up to 15\%, which is of the order of electric potential variations of the ESW. The observed ESWs are responsible for extreme energy conversion rates, up to 2.5 nW/m$^3$. We revealed that ESWs are typically observed within strong, field-aligned Electron-scale Current Sheets (ECSs) generated by suprathermal electron beams accelerated in remote sources. The generation of the observed ESWs could be due to ion/electron acoustic instabilities. We assume that the observed ESWs mark the secondary separatrices of the secondary reconnections and provide an energy transfer from suprathermal to cold electrons in collisionless plasma.

In a three-stroke negative cloud-to-ground flash, the leader of Stroke 2, while forming a new, heavily branched path to ground, briefly collided with the defunct (non-luminous) channel/branches of Stroke 1 at heights of about 2.0 km, 1.6 km, again 2.0 km, and 90 m above ground level. Each of the first three (higher-altitude) collisions was associated with luminosity waves originating from the collision point, including an upward reflection-type wave along the colliding leader channel and one or more transmitted waves along the residual branches of the preceding stroke. Based on the estimated extension speeds and RF field signatures, we attributed the transmitted luminosity waves to dart-stepped leaders developing in relatively short segments of the decayed channel. As a result of the last collision, at a height of 90 m above ground level, a branch of the Stroke 2 leader entered the lower part of the residual (non-luminous) Stroke 1 channel and connected to the ground, producing a return stroke. Additionally, the heavily-branched Stroke 2 leader created a ground termination about 950 m away from its ground termination common with Stroke 1. We observed four X-ray pulses with peaks ranging from 130 to 750 keV during a 3.5-ms or so interval around the time of the higher-altitude collisions during the leader stage of Stroke 2, but no detectable X-ray pulses during the leader stages of Strokes 1 and 3. The probability of random occurrence of 4 X-ray pulses within a 3.5-ms time interval at our site is 0.00099.

Many plant collections of historical and conservation focus are housed within botanic gardens worldwide, serving as important sources of research, conservation, and education. Botanists have long been interested in cycads with their links to early seed plant evolution. In 2024, the Missouri Botanical Garden's extensive cycad collection contains forty-nine taxa and nine of the ten extant genera. Many of these plants have an eclectic history and some of the living accessions can be traced back to the 1904 World's Fair, and several cycads in the living collections at Missouri Botanical Garden could have enhanced conservation value if their provenance can be determined via genomics and morphological comparisons. Plant conservation is increasingly being hailed as a central tenant of the mission of contemporary botanic gardens during the Anthropocene, and engaging in cycad conservation via ex situ conservation, research, and education will serve to forward this critical mission.

Agricultural intensification in Lao PDR has increased significantly each year, resulting in the conversion of vast natural habitats to various levels of agricultural intensification. However, there has been a longstanding knowledge gap regarding the impact of agricultural intensification on native fauna species in the country. The present study was conducted in two different agricultural systems (organic and conventional), which varied in land-use intensity, in central Lao PDR. Within each farming system, five sampling plots (30 x 50 m), spaced 100 m apart, and 900 m distance between organic and conventional farming systems were selected. The farming system had the greatest influence on native honey bee abundance. Although honey bee abundance was higher in the organic, the diversity of flowering plant resources did not differ significantly between the two farming systems. This suggests that the farming system has a more crucial impact on native honey bee populations than on the diversity of flowering plants. In organic farming, it was found that native honey bee populations could provide full pollination services, even for crops with high pollination demands without the need for other pollinators. All other sampling plots in the conventional farming system, however, showed a significant decline in native bee abundance, resulting in insufficient pollination services from native bees alone. This study found that diversity is crucial for sustaining these services, given the year-to-year variation in population levels. Additional measures are needed to conserve more specialized native honey bee species and other pollinators in both farming systems.