NMR Biomed. 2025 Jun;38(6):e70038. doi: 10.1002/nbm.70038.

ABSTRACT

Accurate quantification of metabolites in magnetic resonance spectroscopy (MRS) is challenged by low signal-to-noise ratio (SNR), overlapping metabolites, and various artifacts. Particularly, unknown and unparameterized baseline effects obscure the quantification of low-concentration metabolites, limiting MRS reliability. This paper introduces wavelet analysis-based neural decomposition (WAND), a novel data-driven method designed to decompose MRS signals into their constituent components: metabolite-specific signals, baseline, and artifacts. WAND takes advantage of the enhanced separability of these components within the wavelet domain. The method employs a neural network, specifically a U-Net architecture, trained to predict masks for wavelet coefficients obtained through the continuous wavelet transform. These masks effectively isolate desired signal components in the wavelet domain, which are then inverse-transformed to obtain separated signals. Notably, an artifact mask is created by inverting the sum of all known signal masks, enabling WAND to capture and remove even unpredictable artifacts. The effectiveness of WAND in achieving accurate decomposition is demonstrated through numerical evaluations using simulated spectra. Furthermore, WAND's artifact removal capabilities significantly enhance the quantification accuracy of linear combination model fitting. The method's robustness is further validated using data from the 2016 MRS Fitting Challenge and in vivo experiments.

PMID:40289522 | DOI:10.1002/nbm.70038

Expert Rev Respir Med. 2025 Apr 27. doi: 10.1080/17476348.2025.2499651. Online ahead of print.

ABSTRACT

INTRODUCTION: Idiopathic pleuroparenchymal fibroelastosis (iPPFE) is a rare progressive interstitial lung disease characterized by upper-lobe fibrosis, severe restrictive impairment, and poor prognosis. Unlike idiopathic pulmonary fibrosis, in which acute exacerbations, chronic respiratory failure, and lung cancer are the major causes of death, iPPFE primarily leads to progressive respiratory failure, often complicated by malnutrition and recurrent pneumothorax. Despite growing recognition, its pathogenesis remains unclear and no effective treatments exist.

AREAS COVERED: This review summarizes the epidemiological, clinical, radiological, and pathological features of iPPFE, as well as diagnostic and prognostic advancements. Key prognostic factors include weight loss, reduced forced vital capacity, hypercapnia, and lower-lobe interstitial pneumonia. Serum biomarkers (e.g. latent transforming growth factor-beta binding protein-4) are being explored for early detection and prognostic purposes. Although antifibrotic agents show limited efficacy, supportive care - pulmonary rehabilitation, nutritional management, and pneumothorax prevention - remains essential. Research on the fibroelastotic pathways may inform the development of future therapies.

EXPERT OPINION: IPPFE remains a challenging disease. Therefore, early diagnosis and comprehensive management of this condition are crucial. Future research should refine prognostic models and explore novel therapeutic approaches for treating fibroelastosis. Lung transplantation may be an option for select patients. Further studies are required to optimize these outcomes.

PMID:40289399 | DOI:10.1080/17476348.2025.2499651

Br J Clin Pharmacol. 2025 Apr 27. doi: 10.1002/bcp.70057. Online ahead of print.

ABSTRACT

AIMS: Pharmacogenomics enables personalization of drug therapy improving effectiveness and/or safety. Dihydropyrimidine dehydrogenase [DPYD] testing prior to fluoropyrimidine chemotherapy was commissioned by NHS England in response to an update from the Medicines and Healthcare products Regulatory Agency.

METHODS: A questionnaire developed and approved by the National DPYD Working Group investigated processes of testing, receiving and responding to results. The survey was distributed to Genomics Medicine Service Alliance (GMSA) Clinical Directors and Pharmacy Senior Responsible Officers for dissemination in their geography. Data were collected June-September 2021. All hospitals delivering fluoropyrimidines across the UK were invited to participate.

RESULTS: 131/138 (94.9%) organizations reported testing all patients receiving fluoropyrimidines. In England 76.7% of hospitals sent samples to centrally commissioned genomics laboratories. In the devolved nations, 73.9% sent samples to regional genomics laboratories. Multidisciplinary staff including oncologists, independent non-medical prescribers, clinical nurse specialists, screening pharmacists and chemotherapy nurses requested the test, checked and actioned the result. Self-reported turnaround times varied from <2 days where a regional laboratory was colocated with a chemotherapy centre to >10 days.

CONCLUSION: Through multiprofessional, national collaboration, this is the first report studying the large-scale rollout of a nationally commissioned pharmacogenetic test. Whilst DPYD testing has been successfully implemented, there is a need to standardize and improve end-to-end turnaround times. This has led to the development of a best practice pathway. Critically, GMSAs must build on this implementation to deliver its priorities, in supporting equitable access to future pharmacogenomic testing across a wider cohort of therapies.

PMID:40289265 | DOI:10.1002/bcp.70057

Methods. 2025 Apr 25:S1046-2023(25)00109-4. doi: 10.1016/j.ymeth.2025.04.012. Online ahead of print.

ABSTRACT

Drug-target interaction (DTI) prediction is critical in drug repositioning and discovery. In current metapath-based prediction methods, attention mechanisms are often used to differentiate the importance of various neighbors, enhancing the model's expressiveness. However, in biological networks with small-scale imbalanced data, attention mechanisms are prone to interference from noise and missing data, leading to instability in weight learning, reduced efficiency, and an increased risk of overfitting. To address these issues, we propose the use of average aggregation to mitigate noise, simplify model complexity, and improve stability. Specifically, we introduce a simplified mean aggregation method for DTI prediction. This approach uses average aggregation, effectively reducing noise interference, lowering model complexity, and preventing overfitting, making it especially suitable for current biological networks. Extensive testing on three heterogeneous biological datasets shows that SNADTI outperforms 12 leading methods across two evaluation metrics, significantly reducing training time and validating its effectiveness in DTI prediction. Complexity analysis reveals that our method offers a substantial computational speed advantage over other methods on the same dataset, highlighting its enhanced efficiency. Experimental results demonstrate that SNADTI excels in prediction accuracy, stability, and reproducibility, confirming its practicality and effectiveness in DTI prediction.

PMID:40288620 | DOI:10.1016/j.ymeth.2025.04.012

Farm Hosp. 2025 Apr 26:S1130-6343(25)00009-1. doi: 10.1016/j.farma.2025.02.002. Online ahead of print.

ABSTRACT

OBJECTIVE: Inhaled therapy is essential in cystic fibrosis; however, inhalers have a significant environmental impact due to the greenhouse gases (GHGs) emitted. The environmental impact of a product is estimated by its carbon footprint (CF). Pressurized metered-dose inhalers (pMDIs) have a higher CF than dry powder inhalers (DPIs) and soft mist inhalers (SMIs) due to the incorporation of GHGs. The objectives are to analyze the consumption of inhalers (β2-adrenergic agonist bronchodilators, anticholinergics, and/or corticosteroids) in a cystic fibrosis unit and estimate the generated CF.

METHOD: Retrospective determination (January 2018-December 2023) of consumption and CF (tCO2eq) by type of inhaler was conducted. Consumption and CF trends were evaluated using linear regression.

RESULTS: Annually, 1.529 (1.279-1.613) pMDIs, 1.055 (855-1.333) DPIs, and 28 (20-42) SMIs were dispensed, representing 55.97%, 42.33%, and 1.70%, respectively. A statistically significant positive trend in the consumption of SMIs was observed. The median annual CF was: pMDIs 38.3 (31.2-40.3) tCO2eq, DPIs 0.8 (0.6-0.9) tCO2eq, and SMIs 0.02 (0.02-0.03) tCO2eq, representing 97.86%, 2.04%, and 0.10%, respectively.

CONCLUSIONS: pMDIs were the inhalers with the highest consumption and CF, although their consumption appears to be decreasing, with an increase in the consumption of SMIs.

PMID:40288920 | DOI:10.1016/j.farma.2025.02.002

Microb Pathog. 2025 Apr 25:107628. doi: 10.1016/j.micpath.2025.107628. Online ahead of print.

ABSTRACT

BACKGROUND: Cystic fibrosis (CF) lung disease begins early, and prophylactic antibiotics have been used to prevent Staphylococcus aureus infection. This study examined the lung microbiome in two countries with differing antibiotic practices and its relationship to lung function in young children with CF.

METHODS: A binational, longitudinal, observational study was performed to define the lower airway microbiome in infants with CF. 16S rRNA sequencing was performed using lavage fluid to characterize the lung microbiota in 45 infants with and without prophylactic antibiotic therapy at an average age of approximately 3 months and 14 months. The association between pulmonary function, bacterial community diversities, and taxa was assessed.

RESULTS: Expected CF bacterial genera and non-traditional bacteria, such as Streptococcus, were identified as core taxa. Microbial community shifts were observed in infants who received antibiotic prophylaxis, with lower alpha diversity (ANOVA, P<0.05) and a higher proportion of Streptococcus at the first visit. Beta diversity (FEV0.5z; MiRKAT, P<0.05) and Streptococcus were associated with FEV0.5z (LASSO and linear regression, β<0). Functional annotation suggested that alteration of lung microbiota may be linked to antimicrobial resistance.

CONCLUSIONS: Lung microbial diversity in infants with CF varied between the two countries, particularly during early infancy. A shift in the lung microbiome toward a higher relative abundance of Streptococcus was associated with reduced pulmonary function.

PMID:40288428 | DOI:10.1016/j.micpath.2025.107628

Sleep Med. 2025 Apr 23;131:106534. doi: 10.1016/j.sleep.2025.106534. Online ahead of print.

ABSTRACT

Noninvasive ventilation (NIV) is widely used in children. Only a few devices are life support ventilators. The pressure support (PSV) mode is the most common used mode for home NIV, while assist-control pressure ventilation (PAC) is usually used in patients with abnormal central drive. Patient-ventilator asynchrony (PVA) is common during NIV and may have different causes, such as unintentional leaks, inadequate settings or misunderstanding of the settings. However, PVA may also be due to issues related to the NIV device, which is less common and is challenging. We report here the cases of 5 children with PVA due to trigger issues with a recent life support device.

PMID:40288254 | DOI:10.1016/j.sleep.2025.106534

Sci Rep. 2025 Apr 27;15(1):14732. doi: 10.1038/s41598-025-00115-1.

ABSTRACT

Accurate medium- to long-term runoff forecasting is crucial for flood control, drought resilience, water resources development, and ecological improvement. Traditional statistical methods struggle to utilize multifaceted variable information, leading to lower prediction accuracy. This study introduces two innovative coupled models-SRA-SVR and SRA-MLPR-to enhance runoff prediction by leveraging the strengths of statistical and deep learning approaches. Stepwise Regression Analysis (SRA) was employed to effectively handle high-dimensional data and multicollinearity, ensuring that only the most influential predictive variables were retained. Support Vector Regression (SVR) and Multi-Layer Perceptron Regression (MLPR) were chosen due to their strong adaptability in capturing nonlinear relationships and extracting latent hydrological patterns. The integration of these methods significantly improves prediction accuracy and model stability. By integrating 80 atmospheric circulation indices as teleconnection variables, the models tackle critical challenges such as high-dimensional data, multicollinearity, and nonlinear hydrological dynamics. The Yalong River Basin, characterized by complex hydrological processes and diverse climatic influences, serves as the case study for model validation. The results show that: (1) Compared to baseline single models, the SRA-MLPR model reduced RMSE (from 798.47 to 594.45) by 26% and MAPE (from 34.79 to 22.90%) by 34%, while achieving an NSE (from 0.67 to 0.76) improvement of 13%, particularly excelling in extreme runoff scenarios. (2) The inclusion of teleconnection indices not only enriched the predictive feature set but also improved model stability, with the SRA-MLPR demonstrating enhanced capability in capturing latent nonlinear relationships. (3) A one-month lag in atmospheric circulation indices was identified as the optimal predictor for basin-scale runoff, providing actionable insights into temporal runoff dynamics. (4) To enhance model interpretability, SHAP (SHapley Additive exPlanations) analysis was employed to quantify the contribution of atmospheric circulation indices to runoff predictions, revealing the dominant climate drivers and their nonlinear interactions. The results indicate that the Northern Hemisphere Polar Vortex and the East Asian Trough exert significant control over runoff dynamics, with their influence modulated by large-scale climate oscillations such as the North Atlantic Oscillation (NAO) and Pacific Decadal Oscillation (PDO). (5) The models' scalability is validated through their modular design, allowing seamless adaptation to diverse hydrological contexts. Applications include improved flood forecasting, optimized reservoir operations, and adaptive water resource planning. Furthermore, the study demonstrates the potential of coupled models as generalizable tools for hydrological forecasting in basins with varying climatic and geographic conditions. This study highlights the potential of coupled models as robust and generalizable tools for hydrological forecasting across diverse climatic and geographic conditions. By integrating atmospheric circulation indices, the proposed models enhance runoff prediction accuracy and stability while offering valuable insights for flood prevention, drought mitigation, and adaptive water resource management. These methodological advancements bridge the gap between statistical and deep learning approaches, providing a scalable framework for accurate and interpretable hydrological, climatological, and environmental predictions. Given the escalating challenges brought about by climate change, the findings of this study make contributions to sustainable water management, interpretable decision-making support, and disaster preparedness at a global level.

PMID:40289219 | DOI:10.1038/s41598-025-00115-1

Sci Rep. 2025 Apr 27;15(1):14718. doi: 10.1038/s41598-025-99779-y.

ABSTRACT

Intelligent knowledge about Capsicum annuum L. germplasm could lead to effective management of germplasm. Here, 29 accessions of sweet pepper were investigated in two separate randomized complete block design with three replications in the field condition. Fruit yield accompanied by 13 agro-morphological traits were recorded in two experiments. Genomic fingerprinting of accessions was done by using 10 ISSR primers. The convolutional neural network (CNN) models via outputs of both correlation coefficients and stepwise regression showed the high accuracy of CNN model through correlation coefficients (R2 = 0.879) in predicting fruit yield of sweet pepper. Fruit thickness and fruit width were identified simultaneously as significant components in both models. Genomic best linear unbiased prediction through 65 amplified ISSR loci showed positive and high value of additive gene effect as breeding value for traits identified in the deep learning models. Among studied germplasm, G12, G13, G14, and G25 with positive and high value of breeding value especially for traits constructed the CNN models, recognized as superior genotypes. Regardless of breeding value, multi-trait genotype-ideotype distance index by utilizing all recorded agro-morphological traits simultaneously, revealed G11, G12, G13, and G15 as promising genotypes. So, G12 and G13 which have ideal values of studied traits simultaneously and also positive breeding value could be considered as promising parents for future breeding programs. The study concludes that a CNN model focusing on morphological traits with additive genetic control, combined with the MTSI index, can effectively enhance parental selection in sweet pepper.

PMID:40289216 | DOI:10.1038/s41598-025-99779-y

Sci Rep. 2025 Apr 27;15(1):14733. doi: 10.1038/s41598-025-99777-0.

ABSTRACT

This study aims to effectively improve the quality evaluation system of engineering practice teaching in colleges and universities and enhance the efficiency of teaching management. A brand-new teaching evaluation model is constructed based on the Internet of Things (IoT) technology, combined with complex network analytic hierarchy process and deep learning method. Firstly, with the help of open online course data, Natural Language Processing (NLP) technology and Generative Adversarial Network (GAN) algorithm are used to extract discipline-related features from the course content, and the data of 500 students in 10 majors are simulated and generated. Then, the real university curriculum content, teaching resources, and virtual student data are organically integrated, and two deep learning algorithms, Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN), are introduced. RNN is used to capture time series information, and CNN is used to extract spatial features. Through the hierarchical analysis of complex network, the relationship between different teaching elements is revealed and the hierarchical structure is constructed. Meanwhile, dynamic characteristics are introduced, and continuous model updating and adaptation are realized by randomly combining data to adapt to the changes of actual educational environment. After the course training, data indicators such as students' homework, projects and exams are comprehensively extracted, and the correlation analysis between students' performance and characteristics, time series analysis, feature fusion and weight analysis, model performance evaluation and prediction analysis are carried out. Through the correlation analysis between students' performance and characteristics, the important characteristics that affect learning results are excavated. Time series analysis reveals the changing trend of learning process and better grasps students' learning state. Feature fusion and weight analysis comprehensively consider multiple key features to quantify students' comprehensive performance under different parameter characteristics. Model performance evaluation and prediction analysis compare the prediction results of the model with the actual performance to evaluate the accuracy and stability of the model. The results show that there is a positive correlation between curriculum dependence and interdisciplinary impact index (r = 0.725). The performance of student 3 is relatively stable, with the highest score of 91, and the score of students 7 fluctuates the most, from the lowest 47.9 to the highest 50.2. CNN characteristic index and RNN characteristic index are between 0.18 and 0.78. The comprehensive accuracy of the model in predicting students' actual grades reaches 76-98%, and the prediction consistency varies from 76 to 98%. This study aims to help reveal the relationship between students' performance and teaching evaluation factors, deepen the understanding of the evaluation model of engineering practice teaching in colleges and universities, and provide valuable guidance for optimizing teaching.

PMID:40289170 | DOI:10.1038/s41598-025-99777-0

Orphanet J Rare Dis. 2025 Apr 27;20(1):201. doi: 10.1186/s13023-025-03706-3.

ABSTRACT

BACKGROUND: Data on clinical manifestations of neurofibromatosis-Noonan syndrome (NF-NS) remain heterogeneous, with limited validated descriptions.

METHODS: This study aims to better define the clinical and molecular features of NF-NS and compare them with existing literature. Secondary objectives include evaluating inter-rater diagnostic agreement among experienced clinicians and assessing the utility of deep-learning algorithms (Face2Gene® [F2G]). Additionally, we assess the prevalence of congenital heart malformations (CHM) in NF-NS compared to 'classic' neurofibromatosis type 1 (NF1). A 9-year, prospective, monocentric study was conducted, involving patients with NF1 pathogenic variants (PVs) and Noonan syndrome-like facial phenotype (NSLFP).

RESULTS: Twenty-six patients were enrolled. NSLFP was categorized as 'suggestive' in 69% of cases and 'typical' in 31%. The presence of at least two facial abnormalities (e.g., low-set ears, downslanted palpebral fissures, hypertelorism, and ptosis) was consistently observed in 'typical' cases. Inter-rater concordance was substantial (0.65 [95% CI = 0.56; 0.74]), while concordance between clinicians and F2G was almost perfect at (0.821 [CI 95% = 0.625; 1.000]). Missense NF1 PVs were observed in 38.5% of cases. Apart from NSLP and a high frequency of pectus excavatum (62.5%), no significant differences in anthropometric, dermatological, neurological, skeletal, or ocular clinical features were observed between NF-NS and 'classic' NF1. CHM were found in 19.2% of NF-NS patients, with pulmonic stenosis present in 7.7%.

CONCLUSION: NF-NS is a distinct phenotypic variant of NF1, marked by NSLP with consistent facial features -, and frequent pectus excavatum. F2G demonstrated high diagnostic concordance, reinforcing its clinical utility. Given the elevated risk of CHM, especially pulmonic stenosis, proactive cardiovascular assessment similar to other RASopathies is recommended for NS-NF patients, regardless of NF1 PV type.

PMID:40289159 | DOI:10.1186/s13023-025-03706-3

Sci Rep. 2025 Apr 27;15(1):14714. doi: 10.1038/s41598-025-97109-w.

ABSTRACT

Artificial intelligence allows improvements in renewable energy systems by increasing efficiency while enhancing reliability and reducing costs. Renewable energy forecasting receives substantial improvement by applying deep learning methods as one of its promising approaches. The research utilizes QTM with NiOA optimization for achieving maximum forecasting performance. NiOA functions through critical optimization processes when enhancing deep learning models with high accuracy for large complex datasets by selecting the most appropriate features. Fundamental data preparation steps, including normalization scaling, and gap handling, play a vital role before using input data for reliable renewable energy forecasting operations. Using the Ninja binary optimization engine produces superior results than all tested binary algorithms, including SBO, bSCA, bFA, bGA, bFEP, bGSA, bDE, bTSH and bBA, resulting in enhanced classification accuracy. The superior capability of bNinja to choose optimal features establishes its usefulness for renewable energy forecasting applications. Experimental implementation revealed that incorporating the Ninja Optimization Algorithm with the QTM model delivered the best R2 performance at 95.15% with an exceptional RMSE value of 0.00003, thus establishing its ability to optimize renewable energy forecasting accuracy.

PMID:40289143 | DOI:10.1038/s41598-025-97109-w

J Biophotonics. 2025 Apr 27:e70047. doi: 10.1002/jbio.70047. Online ahead of print.

ABSTRACT

Early identification of gastric adenomatous polyps and adenocarcinoma is vital for improving patient outcomes. This study proposes a hybrid CNN-GRU model to classify one-dimensional hyperspectral data from ex vivo gastric tissues, addressing limitations of traditional diagnostics. Our model innovatively combines convolutional neural networks (CNNs) and gated recurrent units (GRUs) to capture both spatial and sequential dependencies in spectral data. Experimental results demonstrate that our model achieves an accuracy of 86%, sensitivity of 88%, and specificity of 85%. Additionally, receiver operating characteristic analysis further underscores its robust performance with an area under the curve of 0.86, surpassing traditional methods and other baseline models. These findings highlight the potential of leveraging advanced machine learning techniques to enhance early diagnostic accuracy and treatment strategies. The proposed approach offers a promising tool for rapid, accurate differentiation of gastric lesions, underscoring the importance of integrating innovative technologies in clinical diagnostics.

PMID:40288998 | DOI:10.1002/jbio.70047

Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2025 Apr 25;42(2):396-403. doi: 10.7507/1001-5515.202403004.

ABSTRACT

Neurodegenerative diseases (NDDs) are a group of heterogeneous neurological disorders that can cause progressive loss of neurons in the central nervous system or peripheral nervous system, resulting in a decline in motor function. Motion capture, as a high-precision and high-resolution technology for capturing human motion data, drives NDDs motor assessment to effectively extract kinematic features and thus assess the patient's motor ability or disease severity. This paper focuses on the recent research progress in motor assessment of NDDs driven by motion capture data. Based on a brief introduction of NDDs motor assessment datasets, we categorized the assessment methods into three types according to the way of feature extraction and processing: NDDs motor assessment methods based on statistical analysis, machine learning and deep learning. Then, we comparatively analyzed the technical points and characteristics of the three types of methods from the aspects of data composition, data preprocessing, assessment methods, assessment purposes and effects. Finally, we discussed and prospected the development trends of NDDs motor assessment.

PMID:40288984 | DOI:10.7507/1001-5515.202403004

Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2025 Apr 25;42(2):272-279. doi: 10.7507/1001-5515.202411035.

ABSTRACT

Motor imagery (MI) is a mental process that can be recognized by electroencephalography (EEG) without actual movement. It has significant research value and application potential in the field of brain-computer interface (BCI) technology. To address the challenges posed by the non-stationary nature and low signal-to-noise ratio of MI-EEG signals, this study proposed a Riemannian spatial filtering and domain adaptation (RSFDA) method for improving the accuracy and efficiency of cross-session MI-BCI classification tasks. The approach addressed the issue of inconsistent data distribution between source and target domains through a multi-module collaborative framework, which enhanced the generalization capability of cross-session MI-EEG classification models. Comparative experiments were conducted on three public datasets to evaluate RSFDA against eight existing methods in terms of classification accuracy and computational efficiency. The experimental results demonstrated that RSFDA achieved an average classification accuracy of 79.37%, outperforming the state-of-the-art deep learning method Tensor-CSPNet (76.46%) by 2.91% ( P < 0.01). Furthermore, the proposed method showed significantly lower computational costs, requiring only approximately 3 minutes of average training time compared to Tensor-CSPNet's 25 minutes, representing a reduction of 22 minutes. These findings indicate that the RSFDA method demonstrates superior performance in cross-session MI-EEG classification tasks by effectively balancing accuracy and efficiency. However, its applicability in complex transfer learning scenarios remains to be further investigated.

PMID:40288968 | DOI:10.7507/1001-5515.202411035

Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2025 Apr 25;42(2):255-262. doi: 10.7507/1001-5515.202407017.

ABSTRACT

Among numerous medical imaging modalities, diffusion weighted imaging (DWI) is extremely sensitive to acute ischemic stroke lesions, especially small infarcts. However, magnetic resonance imaging is time-consuming and expensive, and it is also prone to interference from metal implants. Therefore, the aim of this study is to design a medical image synthesis method based on generative adversarial network, Stroke-p2pHD, for synthesizing DWI images from computed tomography (CT). Stroke-p2pHD consisted of a generator that effectively fused local image features and global context information (Global_to_Local) and a multi-scale discriminator (M 2Dis). Specifically, in the Global_to_Local generator, a fully convolutional Transformer (FCT) and a local attention module (LAM) were integrated to achieve the synthesis of detailed information such as textures and lesions in DWI images. In the M 2Dis discriminator, a multi-scale convolutional network was adopted to perform the discrimination function of the input images. Meanwhile, an optimization balance with the Global_to_Local generator was ensured and the consistency of features in each layer of the M 2Dis discriminator was constrained. In this study, the public Acute Ischemic Stroke Dataset (AISD) and the acute cerebral infarction dataset from Yantaishan Hospital were used to verify the performance of the Stroke-p2pHD model in synthesizing DWI based on CT. Compared with other methods, the Stroke-p2pHD model showed excellent quantitative results (mean-square error = 0.008, peak signal-to-noise ratio = 23.766, structural similarity = 0.743). At the same time, relevant experimental analyses such as computational efficiency verify that the Stroke-p2pHD model has great potential for clinical applications.

PMID:40288966 | DOI:10.7507/1001-5515.202407017

BMJ Health Care Inform. 2025 Apr 27;32(1):e101130. doi: 10.1136/bmjhci-2024-101130.

ABSTRACT

The recent emergence of foundation model-based chatbots, such as ChatGPT (OpenAI, San Francisco, CA, USA), has showcased remarkable language mastery and intuitive comprehension capabilities. Despite significant efforts to identify and address the near-term risks associated with artificial intelligence (AI), our understanding of the existential threats they pose remains limited. Near-term risks stem from AI that already exist or are under active development with a clear trajectory towards deployment. Existential risks of AI can be an extension of the near-term risks studied by the fairness, accountability, transparency and ethics community, and are characterised by a potential to threaten humanity's long-term potential. In this paper, we delve into the ways AI can give rise to existential harm and explore potential risk mitigation strategies. This involves further investigation of critical domains, including AI alignment, overtrust in AI, AI safety, open-sourcing, the implications of AI to healthcare and the broader societal risks.

PMID:40288807 | DOI:10.1136/bmjhci-2024-101130

BMC Microbiol. 2025 Apr 27;25(1):253. doi: 10.1186/s12866-025-03968-8.

ABSTRACT

BACKGROUND: Research has established links between the gut microbiome (GM) and both obesity and type 2 diabetes (T2D), which is much discussed, but underexplored. This study employed body mass index (BMI) as the measurement of obesity to delve deeper into the correlations from a genetic perspective.

METHODS: We performed the Mendelian randomization (MR) analysis to examine the causal effects of GM on T2D and BMI, and vice versa. Genome-wide association study (GWAS) summary datasets were utilized for the analysis, including T2D (N = 933,970), BMI (N = 806,834), and two GM datasets from the international consortium MiBioGen (211 taxa, N = 18,340) and the Dutch Microbiome Project (DMP) (207 taxa, N = 7,738). These datasets mainly cover European populations, with additional cohorts from Asia and other regions. To further explore the potential mediating role of GM in the connections between BMI and T2D, their interaction patterns were summarized into a network.

RESULTS: MR analysis identified 9 taxa that showed protective properties against T2D. Seven species were within the Firmicutes and Bacteroidales phyla in the DMP, and two were from the MiBioGen (Odds Ratio (OR): 0.94-0.95). Conversely, genetic components contributing to the abundance of 12 taxa were associated with increased risks of T2D (OR: 1.04-1.12). Furthermore, T2D may elevate the abundance of seven taxa (OR: 1.03-1.08) and reduce the abundance of six taxa (OR: 0.93-0.97). In the analysis of the influence of the genetic component of BMI on GM composition, BMI affected 52 bacterial taxa, with 28 decreasing (OR: 0.75-0.92) and 24 increasing (OR: 1.08-1.27). Besides, abundances of 25 taxa were negatively correlated with BMI (OR: 0.95-0.99), while positive correlations were detected for 14 taxa (OR: 1.01-1.05). Notably, we uncovered 11 taxa genetically associated with both BMI and T2D, which formed an interactive network.

CONCLUSIONS: Our findings provide evidence for the GM-mediated links between obesity and T2D. The identification of relevant GM taxa offers valuable insights into the potential role of the microbiome in these diseases.

PMID:40289103 | DOI:10.1186/s12866-025-03968-8

BMC Biol. 2025 Apr 28;23(1):112. doi: 10.1186/s12915-025-02216-9.

ABSTRACT

BACKGROUND: Plant growth and morphogenesis is a mechanical process controlled by genetic and molecular networks. Measuring mechanical properties at various scales is necessary to understand how these processes interact. However, obtaining a device to perform the measurements on plant samples of choice poses technical challenges and is often limited by high cost and availability of specialized components, the adequacy of which needs to be verified. Developing software to control and integrate the different pieces of equipment can be a complex task.

RESULTS: To overcome these challenges, we have developed a computer automated micro-extensometer combined with low-cost optical tracking (Camelot) that facilitates measurements of elasticity, creep, and yield stress. It consists of three primary components: a force sensor with a sample attachment point, an actuator with a second attachment point, and a camera. To monitor force, we use a parallel beam sensor, commonly used in digital weighing scales. To stretch the sample, we use a stepper motor with a screw mechanism moving a stage along linear rail. To monitor sample deformation, a compact digital microscope or a microscope camera is used. The system is controlled by MorphoRobotX, an integrated open-source software environment for mechanical experimentation. We first tested the basic Camelot setup, equipped with a digital microscope to track landmarks on the sample surface. We demonstrate that the system has sufficient accuracy to measure the stiffness in delicate plant samples, the etiolated hypocotyls of Arabidopsis, and were able to measure stiffness differences between wild type and a xyloglucan-deficient mutant. Next, we placed Camelot on an inverted microscope and used a C-mount microscope camera to track displacement of cell junctions. We stretched onion epidermal peels in longitudinal and transverse directions and obtained results similar to those previously published. Finally, we used the setup coupled with an upright confocal microscope and measured anisotropic deformation of individual epidermal cells during stretching of an Arabidopsis leaf.

CONCLUSIONS: The portability and suitability of Camelot for high-resolution optical tracking under a microscope make it an ideal tool for researchers in resource-limited settings or those pursuing exploratory biomechanics work.

PMID:40289087 | DOI:10.1186/s12915-025-02216-9

Eur Urol. 2025 Apr 26:S0302-2838(25)00210-6. doi: 10.1016/j.eururo.2025.04.008. Online ahead of print.

ABSTRACT

BACKGROUND AND OBJECTIVE: Divergent differentiation and histologic subtypes are common findings in urothelial carcinoma (UC). Clinically relevant genomic alterations and oncogenic drivers of individual subtypes remain poorly defined. We characterized surgical outcomes and the genomic landscape of UC with aberrant histology (UCAH), with a focus on biomarkers and targetable alterations.

METHODS: The clinical cohort comprised 3052 patients who underwent radical cystectomy (RC) with or without neoadjuvant chemotherapy. Targeted exon sequencing was performed for a genomic cohort of 1060 bladder tumors from RC or transurethral resection specimens. We characterized the frequency of oncogenic mutations and targetable alterations, and the tumor mutational burden (TMB) of each subtype. We defined the clonal relatedness of morphologically distinct regions of tumors with mixed histology.

KEY FINDINGS AND LIMITATIONS: Patients with plasmacytoid, micropapillary, sarcomatoid, or mixed-histology tumors had worse cancer-specific survival than patients with pure urothelial histology. ERBB2, FGFR3, and PTEN alterations were most frequent in micropapillary, nested/squamous, and sarcomatoid UC, respectively. TMB was highest in plasmacytoid, neuroendocrine, and micropapillary tumors. Regions of mixed histology had shared clonal origins, but exceptions were observed. The retrospective design and potential for selection bias are limitations of our study.

CONCLUSIONS AND CLINICAL IMPLICATIONS: UCAH tumors have distinct patterns of genomic alterations, which may be targetable via novel therapies and have implications for clinical trial inclusion. Biomarker-driven systemic therapy should be explored in patients with histologic subtypes that are associated with worse clinical outcomes.

PMID:40288936 | DOI:10.1016/j.eururo.2025.04.008