Browse Articles

Classification of current density vector map using transformer hybrid residual network

The classification of the current density vector map (CDVM) reconstructed from magnetocardiogram (MCG) is an important indicator for assessing cardiac function and state in clinical diagnosis. Given the limited widespread application of MCG, research on CDVM often encounters challenges such as scarcity of data and difficulties in judgment. There is growing interest in computer-aided methods to assist physicians in analyzing cardiac cases using CDVM. This paper proposes a deep learning-based approachto classify the CDVM. To address the issue of insufficient processed MCG data, data augmentation is carried out by adding noise, making predictions based on auto regressive integrated moving average (ARIMA) model, and utilizing interpolation methods. A transformer hybrid residual network is then employed to classify the CDVM across categories 0 to 4, with transfer learning incorporated into the network structure to initialize model parameters, and the self-attention mechanism of the Transformer enhancing the feature extraction capability. This method achieved a classification accuracy of 97.52%, outperforming previous deep learning approaches, exhibiting both high precision and efficiency. Furthermore, its high scalability ensures that it will continue to meet the evolving needs of physicians as CDVM datasets undergo continuous expansion.

Enhanced YOLOv8 for accurate and efficient floating object detection on water surfaces

Comprehensive analysis of genes associated with necroptosis and pyroptosis in intestinal ischemia-reperfusion injury

Background Intestinal ischemia–reperfusion (II/R) injury is a severe clinical condition in which regulated cell death programs—including pyroptosis and necroptosis—have emerged as key drivers of tissue damage and inflammation. We sought to delineate cell-death–related molecular signatures and candidate therapeutic targets in II/R injury. Methods We obtained transcriptome datasets from Gene Expression Omnibus (GEO) databases for mice (GSE96733, GSE232246) and humans (GSE37013). We cross-referenced genes associated with necroptosis and pyroptosis with differentially expressed genes to identify death-related features. Hub genes were identified through the topological structure of protein interaction networks and validated using an internal validation set, an independent validation set, WGCNA, and qRT-PCR. These genes were also associated with immune cell infiltration. Drug–gene interactions were predicted using DGIdb and verified through molecular docking. Results We identified 1,027 differentially expressed genes (DEGs) in the training set and derived 7 cell death-related differentially expressed genes (DCDEGs) by intersecting gene sets associated with necroptosis and pyroptosis. PPI-based prioritization identified four hub genes— Il1 β, Ripk3, Sting1 (Tmem173), and Tnfaip3 —suggesting cross-regulatory interactions between inflammation and cell death in ischemia-reperfusion pathology. These hub genes were validated using WGCNA analysis and an internal validation set. Immune infiltration analysis indicated significant correlations between hub genes and multiple immune compartments. A predictive model showed good discrimination in the discovery data, and 54 candidate drugs targeting the hub genes were retrieved. qRT-PCR confirmed dysregulation of three hub genes. Conclusion Il1 β, Ripk3, Sting1 , and Tnaip3 were identified as hub genes associated with necroptosis and pyroptosis in intestinal ischemia-reperfusion (II/R) injury. This study provides a reproducible framework and identifies testable targets for translational exploration.

A hybrid machine vision and handcrafted features fusion based approach for fine-grained millet classification

Patient safety risk associated with synchronous telehealth: A scoping review

Objective We aimed to analyze the risks associated with patient safety in synchronous telehealth. Methods Scoping review with search in 3 databases, Medical Literature Analysis and Retrieval System Online (Medline), via PubMed®, Embase® via Elsevier®, and the Cochrane Database of Systematic Reviews, recovering evidence from inception until September 4 th , 2024. Eligible reviews investigated patient safety concerns arising from real-time interactions between healthcare professionals and patients through information and communication technologies (ICT), including telephones and videoconferencing tools. We included systematic reviews examining real-time telehealth interactions between healthcare professionals and patients, addressing safety concerns. We followed standard Joanna Briggs Institute methods for conducting the scoping review. Results A total of 3,641 titles and abstracts were retrieved, and after screening, 15 systematic reviews were included, encompassing 315 studies. These reviews addressed various patient populations, healthcare settings, and telehealth interventions, including virtual consultations, telepharmacy, and telerehabilitation. All 15 reviews reported patient safety risks associated with telehealth, the most frequently reported concern was the patient’s experience, highlighted in 53.3% (n = 8) of the included studies. Additional concerns involved user knowledge gaps and the lack of safety criteria in evaluation protocols. These risks were categorized into five domains: patient experience, safety in prescribing medication, effective communication, training and education, and patient identification. Conclusion This scoping review provides evidence that, although telehealth offers valuable alternatives for healthcare delivery, all evidence highlights specific patient safety risks that require attention. Further research is essential to better understand and mitigate these risks. Strategic investments in education, training, and structured implementation are critical to minimizing adverse events and enhancing patient safety in synchronous telehealth.

Radiomics-based quantification of tumor infiltration in the non-enhancing peritumoral region on postoperative MRI is associated with survival in glioblastoma

Efficacy of pulsed ultraviolet (PUV) light for disinfection of nosocomial pathogens: An in-vitro investigation of key parameters for surface and equipment applications

Hospital-acquired infections remain a persistent challenge, particularly when caused by bacterial strains that have developed resistance to multiple antibiotics. Reports from both clinical wards and public health agencies show how rapidly these organisms adapt, leaving many standard cleaning procedures and treatments far less effective than they once were. This concern prompted us to investigate pulsed ultraviolet (PUV) light from a xenon source as an alternative approach to inactivation, offering rapid action without the use of harsh chemicals. The work involved testing four bacterial species of clinical relevance: Pseudomonas aeruginosa , Staphylococcus aureus , Bacillus cereus , and spores of Bacillus megaterium . We evaluated the impact of UV dose, examined how suspension depth influenced performance, and assessed whether exposure to visible light after treatment could reverse the effect. For all strains, significant microbial reduction was achieved using a broad-spectrum emission with a strong germicidal peak in the 260 nm region. Spores showed much greater resilience than vegetative cells, while increased liquid depth reduced the disinfection efficiency.Some degree of photorepair occurred in non-spore-forming species under standard room lighting. Taken together, these findings indicate that, when tuned to the right parameters, PUV could serve as a valuable addition to hospital disinfection routines, especially for equipment and surfaces that cannot withstand heat or aggressive chemical agents.

Foraminiferal isotopic evidence of abrupt mid-20th century onset of hydrographic instability in Nordic Seas inflow waters

Abstract The flow of warm Atlantic waters into the Nordic Seas largely determines the transport of ocean heat to the Arctic and is a prominent feature of the Atlantic Meridional Overturning Circulation (AMOC). Here we provide a ~ 250-year-long (1750 to 1992 AD), annually- to sub-annually- resolved record of Nordic Seas inflow water characteristics inferred from changes in δ 18 O of planktic foraminiferal carbonate. The new record is reflective of upper ocean temperatures across the Atlantic Water Zone of the Nordic Seas and reveals a previously unrecognized increase in temperature instability ~ AD 1950 that appears to have impacted rates of Greenland Ice Sheet melting, Arctic sea-ice extent and Arctic Surface Air Temperature, delaying the regional response to Anthropogenic Global Warming by several decades. While the relationship between the sudden change in hydrographic conditions and AMOC strength and stability is not yet clear, the change in inflow characteristics ~ AD 1950 was clearly imprinted on deep waters overflowing the Nordic Seas Basin.

Correction: A graph neural network-based approach for predicting SARS-CoV-2–human protein interactions from multiview data

Enhancing the weed segmentation in diverse crop fields using computationally effective concatenated attention U-Net with convolutional block attention module

Abstract Weeds are one of the primary factors that reduce crop productivity by competing for nutrients and water, causing the plant to lose weight and resulting in reduced grain yield. Traditional agricultural practices often rely on uniform herbicide application, which can contaminate soil and raise costs. On agricultural land, selective weed treatment are an efficient and cost-effective way to control weeds that require a deep learning-based crop and weed segmentation system. Many existing crop and weed segmentation research works focus on achieving precise crop and weed segmentation results, rather than building lightweight models to deploy on edge devices. To attain this, we develop an effective and efficient convolutional neural network, namely the Concatenated Attention U-Net with Convolutional Block Attention Module (CAUC). By integrating Linear Concatenated Blocks (LCB), Attention Gate (AG) connections, and Convolutional Block Attention Module (CBAM), the proposed model efficiently utilizes feature maps among its architectural components to achieve superior performance. Depth-wise convolution layers and 1 × 1 convolution layers in LCBs reduce computational complexity. To enable the proposed model to identify the weed portions in multiple crop fields, we integrated three datasets in this research work, namely the Crop/Weed Field Image Dataset (CWFID), Sugar Beet, and Sunflower datasets. Experimental results on carrot, sugar beet, and sunflower crop datasets demonstrate high Accuracy (99.09%), MIoU (81.02%), and F1-score (99.06%), with a modest model size (5.6 MB) and computational parameters (0.377 million). We developed a lightweight computer vision application (13.7 MB) to demonstrate the model’s efficacy on low-computational devices.

Interpretable glucose forecasting for type 2 diabetes across traditional, deep, and large language models

The expression of father-daughter bond behaviors influences adult partner attachment in titi monkeys

Abstract Coppery titi monkeys ( Plecturocebus cupreus ) are socially monogamous monkeys that display strong pair bonds similar to human romantic attachments, preceded by infant attachment to their fathers. To understand how father-daughter bonds impact adult relationship dynamics, we established a novel method for quantifying expression of bond-related behaviors. We assessed behavioral and neural correlates of preference, stress buffering, and separation distress to identify how females’ current and former attachment figures impact female attachment. Whereas all females ( n  = 9) shifted to preferring their partner over father six-months post-pairing, females that exhibited higher expression of juvenile parent preference maintained a relationship with their father six-months post-pairing, as evidenced by higher-than-expected father proximity. Higher expression of juvenile measures of proximity following a brief separation predicted slightly increased partner proximity in adulthood. Neural activity patterns in brain regions assessed pre- and post-pairing showed high similarity in glucose metabolism, despite overall activity being lower post-pairing. While there was some inconsistency in results, higher expression of juvenile proximity following a separation was associated with enhanced reduction in activity within social bonding brain regions (social salience network, periaqueductal gray, cerebellum), suggesting a potential stress buffering benefit via reduced threat-related brain activation, like that seen in high-quality human relationships. These findings advance current knowledge of how early relationships may shape adult bond-related behavior and neural activity.

Garden classification of femoral neck fracture using deep-learning algorithm

Abstract The Garden classification, based on X-ray interpretation and established over 50 years ago, remains the standard clinical classification system for femoral neck fractures (FNFs). Yet, this classification has a high interobserver variability of 70%. We sought to develop a deep-learning algorithm capable of accurately predicting FNF types, using only X-ray images, with performance comparable to that of computed tomography (CT). We retrospectively collected data from 1,588 patients who underwent X-ray and 3D-CT scans and were diagnosed with femoral neck fractures at Asan Medical Center. The input X-ray dataset consisted of paired X-ray images of the hip, with anteroposterior (AP) and lateral views. Using 3D-CT as the reference standard, patients were labeled as Garden types I ( n  = 378, 23.8%), II ( n  = 68, 4.3%), III ( n  = 477, 30.0%), and IV ( n  = 665, 41.9%). Our algorithm consisted of hip-joint detection followed by Garden classification, for which 12 different deep-learning architectures were evaluated. Algorithm performance was externally validated in 100 patients. Our algorithms showed a 90.6% overall accuracy and 88.6% Dice similarity coefficient, indicating excellent FNF type discernment. Our algorithms could serve as a valuable tool for diagnosing FNF based on X-ray data only, with accuracy comparable to that of CT.

Endoscopic ultrasound-guided gastroenterostomy with lumen-apposing metal stent: an animal study comparison of wireless and over-the-wire techniques

Single-cell identification of an endothelial cell proximal SPP1+ macrophage population defines the metastatic vascular niche in lymph nodes

Analysis of anticancer drug associated adverse reactions in depressive patients from vigibase

Bioavailable human metabolites from TOTUM-448 (plant-based formulation) maintain liver cell functionality in a hyperlipidic context that drives MASLD onset

Abstract Lipotoxic and inflammatory environment drives metabolic dysfunction-associated steatotic liver disease (MASLD) onset. As most conventional treatments present adverse side effects, alternative options such as preventive nutritional interventions have been developed, though further clinical validation is needed. In this study, we conducted an innovative ex vivo clinical investigation to examine how circulating metabolites generated after oral intake of TOTUM-448 (a plant-based, polyphenol-rich formulation) may influence hepatocyte function. UHPLC-MS/MS analysis confirmed and characterized the bioavailable polyphenol metabolites present in human serum. This metabolite-enriched serum was further used to treat HepG2 hepatocytes, with or without palmitate pretreatment (250 µM). The effects of TOTUM-448–derived metabolites on hepatocytes were evaluated by monitoring cell viability, lipid metabolism, inflammation, oxidative stress, and endoplasmic reticulum (ER) stress, all of which are central features of MASLD. Treated hepatocytes exhibited resistance to palmitate-induced lipotoxic stress, showing reduced intracellular lipid accumulation. TOTUM-448–derived metabolites also prevented the palmitate-induced upregulation of inflammatory gene expression. Additionally, while palmitate strongly upregulated CHOP and XBP1 mRNA expression as well as ATF6 and Caspase-3 activities, the presence of TOTUM-448–derived metabolites restored these ER stress markers to normal levels.

Impact of obstructive sleep apnea on functional performance and muscle quality of patients with COPD

The impacts of China’s low-carbon technology trade on modern energy access in Africa: the role of domestic absorptive capacity

Unravelling TPX2-centered co-expression networks as key drivers of aggressive prostate cancer

Abstract Prostate cancer (PCa) progression is driven by complex molecular reprogramming, yet distinguishing indolent from aggressive disease remains a challenge. We performed an integrative transcriptomic analysis of 1232 PCa samples spanning normal prostate and all major disease stages including primary localized tumors, metastatic hormone-sensitive PCa (mHSPC), and metastatic castration-resistant PCa (mCRPC). By integrating unsupervised consensus clustering (ATC:hclust), weighted gene co-expression network analysis (WGCNA), and explainable machine learning (ML), we identified key transcriptional programs and biomarkers associated with cancer initiation and disease progression. Our analysis revealed persistent dysregulation of mitotic control, DNA damage repair, transcriptional regulation, and cytoskeletal remodeling, underscoring their functional relevance for PCa progression. We uncovered TPX2 as a central hub gene, consistently upregulated across all disease stages and co-expressed with 21 commonly upregulated genes. ML-based gene ranking and interaction analysis identified connections among the commonly upregulated genes, highlighting CENPA-MYBL2 for primary localized PCa, EXO1-NEIL3 for mHSPC and CENPA-RRM2 for mCRPC. Stage-specific analysis further identified key drivers of distinct disease transitions including EZH2 and PLK1 as major regulators of androgen dependence in mHSPC, and TERT as a hallmark of mCRPC, highlighting its role in telomere maintenance and tumor progression. This study demonstrates that unsupervised clustering combined with WGCNA and ML enables the discovery of clinically relevant molecular signatures in PCa. Our findings establish TPX2 -centered networks together with biological pathways implicated in mitotic regulation and DNA damage repair as key drivers of tumor evolution, providing a biologically informed source for biomarker development, drug testing and mechanistic studies.