ISME J. 2025 Feb 2:wraf018. doi: 10.1093/ismejo/wraf018. Online ahead of print.

ABSTRACT

Microbes transform their environments using diverse enzymatic reactions. However, it remains challenging to measure microbial reaction rates in natural environments. Despite advances in global quantification of enzyme abundances, the individual relationships between enzyme abundances and their reaction rates have not been systematically examined. Using matched proteomic and reaction rate data from microbial cultures, we show that enzyme abundance is often insufficient to predict its corresponding reaction rate. However, we discovered that global proteomic measurements can be used to make accurate rate predictions of individual reaction rates (median R2 = 0.78). Accurate rate predictions required only a small number of proteins and they did not need explicit prior mechanistic knowledge or environmental context. These results indicate that proteomes are encoders of cellular reaction rates, potentially enabling proteomic measurements in situ to estimate the rates of microbially mediated reactions in natural systems.

PMID:39893571 | DOI:10.1093/ismejo/wraf018

Network. 2025 Feb 1:1-40. doi: 10.1080/0954898X.2025.2452280. Online ahead of print.

ABSTRACT

Stroke remains a leading global health concern and early diagnosis and accurate identification of stroke lesions are essential for improving treatment outcomes and reducing long-term disabilities. Computed Tomography (CT) imaging is widely used in clinical settings for diagnosing stroke, assessing lesion size, and determining the severity. However, the accurate segmentation and early detection of stroke lesions in CT images remain challenging. Thus, a Jaccard_Residual SqueezeNet is proposed for predicting stroke from CT images with the integration of the Internet of Things (IoT). The Jaccard_Residual SqueezeNet is the integration of the Jaccard index in Residual SqueezeNet. Firstly, the brain CT image is routed to the Base Station (BS) using the Fractional Jellyfish Search Pelican Optimization Algorithm (FJSPOA) and preprocessing is accomplished by median filter. Then, the skull segmentation is accomplished by ENet and then feature extraction is done. Lastly, Stroke is detected using the Jaccard_Residual SqueezeNet. The values of throughput, energy, distance, trust, and delay determined in terms of routing are 72.172 Mbps, 0.580J, 22.243 m, 0.915, and 0.083S. Also, the accuracy, sensitivity, precision, and F1-score for stroke detection are 0.902, 0.896, 0.916, and 0.906. These findings suggest that Jaccard_Residual SqueezeNet offers a robust and efficient platform for stroke detection.

PMID:39893512 | DOI:10.1080/0954898X.2025.2452280

Expert Opin Drug Saf. 2025 Feb 1. doi: 10.1080/14740338.2025.2461200. Online ahead of print.

ABSTRACT

BACKGROUND: This study aims to analyze adverse events (AEs)associated with naltrexone based on the FAERS database, providing a foundationfor its safety monitoring.

RESEARCH DESIGN AND METHODS: Disproportionality analysis methods,including Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR),Bayesian Confidence Propagation Neural Network (BCPNN), and Multi-Item GammaPoisson Shrinker (MGPS) algorithms, were employed to quantify signals ofnaltrexone-related AEs.

RESULTS: AEs related to naltrexone from the first quarter of 2013 to the fourth quarter of2023 were extracted from the FAERS database for detailed analysis. Among atotal of 41,757,311 reports 28,745 were directly associated with naltrexone,involving 27 organ systems. We identified 110 positive signals for AEsat the preferred term (PT) level using disproportionality analysis,whichincluded known AEs such as agitation, depressed mood, sleep disorder,tremor, delirium, and decreased libido. Additionally, our findings suggestedpotential risks of restless legs syndrome, eosinophilic pneumonia, andotolithiasis, which were not mentioned in the drug's label, therebysupplementing the existing safety information.

CONCLUSIONS: The analysis of the FAERS database identified AEsassociated with naltrexone, contributing to the awareness of clinicalpractitioners and pharmacists regarding the drug-related risk signals.Thisawareness facilitates timely preventive and therapeutic measures, ensuringpatient safety.

PMID:39893547 | DOI:10.1080/14740338.2025.2461200

Med Image Anal. 2025 Jan 29;101:103462. doi: 10.1016/j.media.2025.103462. Online ahead of print.

ABSTRACT

Resting-state functional magnetic resonance imaging (rs-fMRI) is a noninvasive technique pivotal for understanding human neural mechanisms of intricate cognitive processes. Most rs-fMRI studies compute a single static functional connectivity matrix across brain regions of interest, or dynamic functional connectivity matrices with a sliding window approach. These approaches are at risk of oversimplifying brain dynamics and lack proper consideration of the goal at hand. While deep learning has gained substantial popularity for modeling complex relational data, its application to uncovering the spatiotemporal dynamics of the brain is still limited. In this study we propose a novel interpretable deep learning framework that learns goal-specific functional connectivity matrix directly from time series and employs a specialized graph neural network for the final classification. Our model, DSAM, leverages temporal causal convolutional networks to capture the temporal dynamics in both low- and high-level feature representations, a temporal attention unit to identify important time points, a self-attention unit to construct the goal-specific connectivity matrix, and a novel variant of graph neural network to capture the spatial dynamics for downstream classification. To validate our approach, we conducted experiments on the Human Connectome Project dataset with 1075 samples to build and interpret the model for the classification of sex group, and the Adolescent Brain Cognitive Development Dataset with 8520 samples for independent testing. Compared our proposed framework with other state-of-art models, results suggested this novel approach goes beyond the assumption of a fixed connectivity matrix, and provides evidence of goal-specific brain connectivity patterns, which opens up potential to gain deeper insights into how the human brain adapts its functional connectivity specific to the task at hand. Our implementation can be found on https://github.com/bishalth01/DSAM.

PMID:39892220 | DOI:10.1016/j.media.2025.103462

Radiography (Lond). 2025 Jan 31;31(2):102878. doi: 10.1016/j.radi.2025.01.013. Online ahead of print.

ABSTRACT

INTRODUCTION: AI-based segmentation techniques in brain MRI have revolutionized neuroimaging by enhancing the accuracy and efficiency of brain structure analysis. These techniques are pivotal for diagnosing neurodegenerative diseases, classifying psychiatric conditions, and predicting brain age. This scoping review synthesizes current methodologies, identifies key trends, and highlights gaps in the use of automatic and semi-automatic segmentation tools in brain MRI, particularly focusing on their application to healthy populations and clinical utility.

METHODS: A scoping review was conducted following Arksey and O'Malley's framework and PRISMA-ScR guidelines. A comprehensive search was performed across six databases for studies published between 2014 and 2024. Studies focused on AI-based brain segmentation in healthy populations, and patients with neurodegenerative diseases, and psychiatric disorders were included, while reviews, case series, and studies without human participants were excluded.

RESULTS: Thirty-two studies were included, employing various segmentation tools and AI models such as convolutional neural networks for segmenting gray matter, white matter, cerebrospinal fluid, and pathological regions. FreeSurfer, which utilizes algorithmic techniques, are also commonly used for automated segmentation. AI models demonstrated high accuracy in brain age prediction, neurodegenerative disease classification, and psychiatric disorder subtyping. Longitudinal studies tracked disease progression, while multimodal approaches integrating MRI with fMRI and PET enhanced diagnostic precision.

CONCLUSION: AI-based segmentation techniques provide scalable solutions for neuroimaging, advancing personalized brain health strategies and supporting early diagnosis of neurological and psychiatric conditions. However, challenges related to standardization, generalizability, and ethical considerations remain.

IMPLICATIONS FOR PRACTICE: The integration of AI tools and algorithm-based methods into clinical workflows can enhance diagnostic accuracy and efficiency, but greater focus on model interpretability, standardization of imaging protocols, and patient consent processes is needed to ensure responsible adoption in practice.

PMID:39892049 | DOI:10.1016/j.radi.2025.01.013

Soc Sci Med. 2025 Jan 10;367:117690. doi: 10.1016/j.socscimed.2025.117690. Online ahead of print.

ABSTRACT

Alcohol is a highly addictive substance, presenting significant global public health concerns, particularly among adolescents. Previous studies have been limited by traditional research methods, making it challenging to encompass diverse risk factors and automate screening or prediction of adolescents' alcohol use. This study aimed to develop prediction algorithms for adolescent alcohol use in South Korea using machine learning (ML) and deep learning (DL) models, and to identify important features. The study utilized a combination of DL (i.e., Auto-encoder) and ML (i.e., Logistic regression, Ridge, LASSO, Elasticnet, Decision tree, Random forest, AdaBoost, and XGBoost) algorithms to develop the prediction models. It involves 41,239 Korean adolescents and 46 socio-ecological input variables based on cross-sectional data. The analysis revealed that the prediction algorithms had AUC scores ranging from 0.6325 to 0.7214. The feature importance analysis indicates that variables within the domains of sociodemographic characteristics, physical and mental health, behavioral problems, family factors, school factors, and social factors all play significant roles. The developed algorithms enable automatic and early identification of adolescent alcohol use within public health practice settings. By leveraging a comprehensive array of input variables, these methods surpass the limitations of traditional regression approaches, offering novel insights into the critical risk factors associated with alcohol use among Korean adolescents, thereby facilitating early and targeted prevention efforts.

PMID:39892039 | DOI:10.1016/j.socscimed.2025.117690

Int J Med Inform. 2025 Jan 30;196:105812. doi: 10.1016/j.ijmedinf.2025.105812. Online ahead of print.

ABSTRACT

BACKGROUND: With advancements in medical technology and science, chronic obstructive pulmonary disease (COPD), one of the world's three major chronic diseases, has seen numerous remarkable outcomes when combined with artificial intelligence, particularly in disease diagnosis. However, the diagnostic performance of these AI models still lacks comprehensive evidence. Therefore, this study quantitatively analyzed the diagnostic performance of AI models in CT images of COPD patients, aiming to promote the development of related research in the future.

METHODS: PubMed, Cochrane Library, Web of Science, and Embase were retrieved up to September 1, 2024. The QUADAS-2 evaluation tool was used to assess the quality of the included studies. Meta-analysis of the included researches was performed using Stata18, RevMan 5.4, and Meta-Disc 1.4 software to merge sensitivity, specificity and plot a summary receiver operating characteristic curve (SROC). Heterogeneity was assessed using the Q statistic, and sources of inter-study heterogeneity were explored through meta-regression analysis.

RESULTS: Twenty-two of 3280 identified studies were eligible. Meta-analysis was performed on 15 of these studies, encompassing a total of 22,817 patients for which statistical metrics were reported or could be calculated. Seven studies were based on deep learning (DL) model, three on machine learning (ML) model, and five on DL model with multiple-instance learning (MIL) mechanisms. One study evaluated both DL and ML models. The meta-analysis results showed that the pooled sensitivity of all DL and ML models was 86 % (95 %CI 78-91 %), specificity was 87 % (95 %CI 83-91 %), and area under the curve was 93 % (95 %CI 90-95 %). Subgroup analyses revealed no significant difference in diagnostic sensitivity and specificity between DL and ML models (sensitivity 82 % (95 %CI 76-87 %), 93 % (95 %CI 85-97 %); specificity 87 % (95 %CI 79-91 %), 84 % (95 %CI 79-88 %), and the DL model with MIL (sensitivity 87 % (95 %CI 61-96 %); specificity 89 % (95 %CI 78-95 %) improved the performance of DL model, but this improvement was not statistically significant (p > 0.05).

CONCLUSION: Both DL and ML models for diagnosing COPD using CT images exhibited high accuracy. There was no significant difference in diagnostic efficacy between the two types of AI models, and the addition of the MIL mechanism may enhance the performance of the DL model.

PMID:39891985 | DOI:10.1016/j.ijmedinf.2025.105812

Commun Biol. 2025 Feb 1;8(1):155. doi: 10.1038/s42003-025-07599-7.

ABSTRACT

Bambuterol is a long-acting anti-asthmatic prodrug which releases terbutaline. Terbutaline is an agonist of the β2-adrenergic receptors which is formed by decarbamoylation of bambuterol by butyrylcholinesterase. Inhibition of the latter, as well as activation of β2-AR, are of interest for the treatment of Alzheimer's disease (AD). Combining these two activities, bambuterol could express a good clinical efficacy against AD. The present work firstly confirmed the capacity of bambuterol to display in cellulo neuroprotective activities, reduction of Tau hyperphosphorylation and preservation of synapses in rat hippocampal neuronal cultures intoxicated with Aβ peptides. Further, bambuterol, in the form of a liposomal gel, showed a good bioavailability in CNS after intranasal administration, which should reduce any side effects linked to peripheral terbutaline release. Indeed, even if the latter is more selective than other β2-mimetics towards bronchial β2-AR, cardiovascular effects (tachycardia, arrhythmias…) could occur upon cardiac β1-AR activation. Finally, intranasal administration of low doses of bambuterol gel in mice intoxicated with Aβ peptides, prevented long-term spatial memory impairment and showed beneficial effects on the survival of neurons and on synapse preservation.

PMID:39893320 | DOI:10.1038/s42003-025-07599-7

J Pharmacol Exp Ther. 2025 Jan;392(1):100013. doi: 10.1124/jpet.124.002152. Epub 2024 Nov 22.

ABSTRACT

Metformin's potential in treating ischemic stroke and neurodegenerative conditions is of growing interest. Yet, the absence of established systemic and brain pharmacokinetic (PK) parameters at relevant preclinical doses presents a significant knowledge gap. This study highlights these PK parameters and the importance of using pharmacologically relevant preclinical doses to study pharmacodynamics in stroke and related neurodegenerative diseases. A liquid chromatography with tandem mass spectrometry method to measure metformin levels in plasma, brain, and cerebrospinal fluid was developed and validated. In vitro assays examined brain tissue binding and metabolic stability. Intravenous bolus administration of metformin to C57BL6 mice covered a low- to high-dose range maintaining pharmacological relevance. Quantification of metformin in the brain was used to assess brain PK parameters, such as unidirectional blood-to-brain constant (Kin) and unbound brain-to-plasma ratio (Kp, uu, brain). Metformin exhibited no binding in the mouse plasma and brain and remained metabolically stable. It rapidly entered the brain, reaching detectable levels in as little as 5 minutes. A Kin value of 1.87 ± 0.27 μL/g/min was obtained. As the dose increased, Kp, uu, brain showed decreased value, implying saturation, but this did not affect an increase in absolute brain concentrations. Metformin was quantifiable in the cerebrospinal fluid at 30 minutes but decreased over time, with concentrations lower than those in the brain across all doses. Our findings emphasize the importance of metformin dose selection based on PK parameters for preclinical pharmacological studies. We anticipate further investigations focusing on PKs and pharmacodynamics in disease conditions, such as stroke. SIGNIFICANCE STATEMENT: The study establishes crucial pharmacokinetic parameters of metformin for treating ischemic stroke and neurodegenerative diseases, addressing a significant knowledge gap. It further emphasizes the importance of selecting pharmacologically relevant preclinical doses. The findings highlight metformin's rapid brain entry, minimal binding, and metabolic stability. The necessity of considering pharmacokinetic parameters in preclinical studies provides a foundation for future investigations into metformin's efficacy for neurodegenerative disease(s).

PMID:39893000 | DOI:10.1124/jpet.124.002152

BMC Infect Dis. 2025 Feb 1;25(1):153. doi: 10.1186/s12879-025-10538-w.

ABSTRACT

BACKGROUND: Anti-infective medicines are crucial for treating infections, but improper dosing can cause toxicity, resistance and treatment failure. Pharmacogenomics can address genetic variations affecting drug efficacy and safety. Despite the high burden of diseases like TB and HIV in Sri Lanka and South Asia, pharmacogenomic data for these populations are limited. This study aims to fill this gap by investigating pharmacogenomic variants in a South Asian population from Sri Lankan.

METHODS: Pharmacogenomic data on anti-infective medicines were obtained from the PharmGKB database, selecting variants with evidence levels 1 A, 1B, 2 A, and 2B. Sri Lankan genetic data were sourced from an anonymized database of 670 Sri Lankans maintained by the Centre for Genetics and Genomics, Faculty of Medicine, University of Colombo. MAFs were compared between Sri Lankan sub-populations and global data from gnomAD, with statistical significance set at p < 0.05.

RESULTS: MAFs of NAT2 gene rs1041983 and rs1799931 variants were, 43.7% (95%CI:41.1-46.4), 7.3% (95%CI:6.0-8.8), respectively. The UGT1A1 rs4148323 variant had a MAF of 3.5% (95%CI:2.6-4.6). In the CYP2B6 gene, 109 individuals were homozygous for the rs3745274 (poor metaboliser) variant, with a MAF of 39.6% (95%CI:37.0-42.3), while the rs34097093 and rs28399499 variants had no individuals homozygous for the variant (MAF: 0.2% [95%CI:0-0.5] (poor/intermediate metaboliser), and 0.1% [95%CI:0-0.4] (poor/intermediate metaboliser), respectively). The MAFs of the CYP2C19 rs12769205 (poor/intermediate metaboliser), rs4244285 (poor/intermediate metaboliser), rs3758581 (poor/intermediate metaboliser), and rs4986893 (poor/intermediate metaboliser) variants were 41.9% (95%CI:39.3-44.6), 41.9% (95%CI:39.2-44.7), 9.7% (95%CI:8.2-11.4), and 0.5% [(95%CI:0.2-1.1), respectively. Most variants showed significant differences compared to global populations, with some exhibiting higher frequencies, particularly when compared to Europeans. CYP2C19 rs12769205 and rs4244285 exhibited higher MAFs in Sri Lankans compared to both other South Asians and Europeans. The NAT2 rs1041983, NAT2 rs1799931, CYP2C19 rs4986893, CYP2C19 rs3758581, and CYP2B6 rs3745274 variants demonstrated significantly higher MAFs than in Europeans but not significantly different from South Asians.

CONCLUSION: This preliminary study identifies variants in NAT2, UGT1A1, CYP2B6, and CYP2C19 genes relevant to the metabolism of anti-TB drugs, antiretrovirals, and voriconazole among Sri Lankans. Several variants, including CYP2C19 rs12769205 and rs4244285, showed higher MAFs, particularly in comparison to European populations, indicating potential differences in drug response. However, the nature of the study limits the ability to explore clinical correlations with the genotypes, therefore further research focusing on clinical correlation and functional validation is required.

PMID:39893405 | DOI:10.1186/s12879-025-10538-w

Cytokine. 2025 Jan 31;187:156870. doi: 10.1016/j.cyto.2025.156870. Online ahead of print.

ABSTRACT

OBJECTIVE: Preeclampsia, characterized by hypertension and proteinuria, is a medical condition associated with maternal and fetal morbidity and mortality. Previous studies reported conflicting correlations between the eNOS rs2070744 variant and the occurrence of preeclampsia. Due to inconsistencies in findings, the purpose of the present meta-analysis was to explore the precise link between the eNOS rs2070744 variant and the development of preeclampsia.

METHODS: The articles were retrieved from various online sources, including Cochrane Library, Google Scholar, EMBASE, PubMed, and Web of Science databases up to February 2024. Data were analyzed by Review Manager (RevMan) 5.4. We adhered to the PRISMA 2020 guidelines to conduct this meta-analysis.

RESULTS: A total of 26 articles containing 3741 cases and 4920 controls were included for qualitative and quantitative data synthesis. In the overall population, we found a strong correlation between the eNOS rs2070744 variant and higher preeclampsia risk in recessive (CC vs. CT + TT: OR = 1.31, p = 0.017) dominant (CC + CT vs. TT: OR = 1.14, p = 0.051), co-dominant 2 (CC vs. TT: OR = 1.37, p = 0.011) and allelic (C vs. T: OR = 1.14, p = 0.022) models. Our study also explored similar outcomes among the Caucasian population in dominant (CC + CT vs. TT: OR = 1.16, p = 0.048), recessive (CC vs. CT + TT: OR = 1.46, p = 0.027), allele (C vs. T: OR = 1.18, p = 0.044), co-dominant 2 (CC vs. TT: OR = 1.53, p = 0.027), and co-dominant 3 (CC vs. CT: OR = 1.46, p = 0.002) models. Besides, a significant risk of preeclampsia in the African population was observed in co-dominant 2 (CC vs. TT: OR = 2.11, p = 0.009), dominant (CC + CT vs. TT: OR = 1.58, p = 0.002) and allelic (C vs. T: OR = 1.45, p = 0.001) models. However, no association of this polymorphism with preeclampsia risk was reported in Asian and mixed populations.

CONCLUSION: This study suggests a significant correlation between eNOS rs2070744 polymorphism and preeclampsia. However, more research on various ethnic groups is necessary to confirm the association.

PMID:39892025 | DOI:10.1016/j.cyto.2025.156870

Eur J Paediatr Neurol. 2025 Jan 11;54:130-139. doi: 10.1016/j.ejpn.2025.01.004. Online ahead of print.

ABSTRACT

BACKGROUND: Duchenne muscular dystrophy (DMD) is a rare neuromuscular disorder characterized by a progressive decline in muscle function, leading to loss of ambulation, respiratory and cardiac failure, and ultimately death. Improvements in DMD management have increased patient life expectancy; therefore, there is a growing requirement for patients to transfer from paediatric to adult care services. There is also a need for clear recommendations to guide this process.

AIM: To establish international consensus guidelines regarding best practices for transitioning patients with DMD from paediatric to adult care and ensuring continuity of treatment.

METHODS: Consensus statements were developed using the Delphi process and scored using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system. The initiative was led by a steering committee (one non-voting chair and two voting members) who recruited 15 expert panellists to form the consensus group. Following an initial systematic literature search, the consensus group voted in three voting rounds. Round 1 (free-text responses to questions) and Round 2 (importance ranking of statements) were completed using an online survey. Round 3 (voting on final consensus statements) took place during a virtual consensus meeting.

CONSENSUS STATEMENTS: Consensus was reached on 48 statements covering the topics of transition planning, the transition process, post-transfer management, communicating with young people with DMD and supporting them with the transition to adult life.

CONCLUSION: These consensus statements provide guidelines for improving transition practices for young people with DMD and promoting continued care at a comparable standard in adulthood.

PMID:39892019 | DOI:10.1016/j.ejpn.2025.01.004

Sci Rep. 2025 Feb 1;15(1):4021. doi: 10.1038/s41598-025-86251-0.

ABSTRACT

This paper provides a novel approach to estimating CO₂ emissions with high precision using machine learning based on DPRNNs with NiOA. The data preparation used in the present methodology involves sophisticated stages such as Principal Component Analysis (PCA) as well as Blind Source Separation (BSS) to reduce noise as well as to improve feature selection. This purified input dataset is used in the DPRNNs model, where both short and long-term temporal dependencies in the data are captured well. NiOA is utilized to tune those parameters; as a result, the prediction accuracy is quite spectacular. Experimental results also demonstrate that the proposed NiOA-DPRNNs framework gets the highest value of R2 (0.9736), lowest error rates and fitness values than other existing models and optimization methods. From the Wilcoxon and ANOVA analyses, one can approve the specificity and consistency of the findings. Liebert and Ruple firmly rethink this rather simple output as a robust theoretic and empirical framework for evaluating and projecting CO2 emissions; they also view it as a helpful guide for policymakers fighting global warming. Further study can build up this theory to include other greenhouse gases and create methods enabling instantaneous tracking for sophisticated and responsive approaches.

PMID:39893234 | DOI:10.1038/s41598-025-86251-0

Sci Rep. 2025 Feb 1;15(1):3985. doi: 10.1038/s41598-025-88199-7.

ABSTRACT

Biochemical recurrence (BCR) of prostate cancer (PCa) negatively impacts patients' post-surgery quality of life, and the traditional predictive models have shown limited accuracy. This study develops an AI-based prognostic model using deep learning that incorporates androgen receptor (AR) regional features from whole-slide images (WSIs). Data from 545 patients across two centres are used for training and validation. The model showed strong performances, with high accuracy in identifying regions with high AR expression and BCR prediction. This AI model may help identify high-risk patients, aiding in better treatment strategies, particularly in underdeveloped areas.

PMID:39893198 | DOI:10.1038/s41598-025-88199-7

Sci Data. 2025 Feb 1;12(1):196. doi: 10.1038/s41597-025-04382-5.

ABSTRACT

The remarkable progress of deep learning in dermatological tasks has brought us closer to achieving diagnostic accuracies comparable to those of human experts. However, while large datasets play a crucial role in the development of reliable deep neural network models, the quality of data therein and their correct usage are of paramount importance. Several factors can impact data quality, such as the presence of duplicates, data leakage across train-test partitions, mislabeled images, and the absence of a well-defined test partition. In this paper, we conduct meticulous analyses of three popular dermatological image datasets: DermaMNIST, its source HAM10000, and Fitzpatrick17k, uncovering these data quality issues, measure the effects of these problems on the benchmark results, and propose corrections to the datasets. Besides ensuring the reproducibility of our analysis, by making our analysis pipeline and the accompanying code publicly available, we aim to encourage similar explorations and to facilitate the identification and addressing of potential data quality issues in other large datasets.

PMID:39893183 | DOI:10.1038/s41597-025-04382-5

Adv Sci (Weinh). 2025 Feb 1:e2407825. doi: 10.1002/advs.202407825. Online ahead of print.

ABSTRACT

Designing voxelized composite structures via 4D printing involves creating voxel units with distinct material properties that transform in response to stimuli; however, optimally distributing these properties to achieve specific target shapes remains a significant challenge. This study introduces an optimization method combining deep learning (DL) and an evolutionary algorithm, focusing on a solvent-responsive hydrogel as the target material. A sequence-enhanced parallel convolutional neural network is developed and generated a dataset through finite element simulations. This DL model enables high-precision prediction of hydrogel deformation. Furthermore, a progressive evolutionary algorithm (PEA) is proposed by integrating the DL model to construct a DL-PEA framework. This framework supports rapid reverse engineering of the desired shape, and the average design time for specified target shapes is reduced to ≈3.04 s. The present findings illustrate how 4D printing of optimized hydrogel designs can effectively transform in response to environmental stimuli. This work provides a new perspective on the application of hydrogels in 4D printing and presents an efficient tool for optimizing 4D-printed voxelized composite structures.

PMID:39893044 | DOI:10.1002/advs.202407825

Am J Transplant. 2025 Jan 30:S1600-6135(25)00047-4. doi: 10.1016/j.ajt.2025.01.041. Online ahead of print.

ABSTRACT

This study aimed to identify the parenchymal structural features by both CT and histology that associate with death-censored graft failure in recipients of living donor kidneys. We analyzed kidney recipients of ABO-compatible living donor kidneys from 2000-2020 with follow-up through 2023. Cortical volume and thickness, individual medullary pyramid volume and count, glomerular volume, nephrosclerosis, and nephron number were assessed by deep learning models applied to the predonation CT and by morphometric histology analysis from the biopsy at the time of transplantation. There were 3098 recipients followed a median 5 years with 346 graft failure events. In adjusted analyses, the only structural measures associated with graft failure were fewer medullary pyramids on CT and a higher fraction of interstitial fibrosis and tubular atrophy (IFTA) on histology. Having ≤15 pyramids donated occurred in 9% and was associated with a graft failure incidence of 2.5 per 100 person-years compared to 1.6 per 100 person-years in the 17% with ≥26 pyramids donated. Fewer medullary pyramids were associated with a lower 1-year eGFR, which mediated the subsequent risk of graft failure. IFTA >1% is also associated with graft failure. Medullary pyramid count is a potentially useful predonation prognostic biomarker for graft failure in transplant recipients.

PMID:39892790 | DOI:10.1016/j.ajt.2025.01.041

J Dent. 2025 Jan 30:105601. doi: 10.1016/j.jdent.2025.105601. Online ahead of print.

ABSTRACT

OBJECTIVES: This study aimed to develop a deep learning (DL) model for the predictive esthetic evaluation of single-implant treatments in the esthetic zone.

METHODS: A total of 226 samples, each comprising three intraoral photographs and 12 clinical features, were collected for proof of concept. Labels were determined by a prosthodontic specialist using the pink esthetic score (PES). A DL model was developed to predict PES based on input images and clinical data. The performance was assessed and compared with that of two other models.

RESULTS: The DL model achieved an average mean absolute error (MAE) of 1.3597, average root mean squared error (MSE) of 1.8324, a Pearson correlation of 0.6326, and accuracies of 65.93% and 85.84% for differences between predicted and ground truth values no larger than 1 and 2, respectively. An ablation study demonstrated that incorporating all input features yielded the best performance, with the proposed model outperforming comparison models.

CONCLUSIONS: DL demonstrates potential for providing acceptable preoperative PES predictions for single implant-supported prostheses in the esthetic zone. Ongoing efforts to collect additional samples and clinical features aim to further enhance the model's performance.

CLINICAL SIGNIFICANCE: The DL model supports dentists in predicting esthetic outcomes and making informed treatment decisions before implant placement. It offers a valuable reference for inexperienced and general dentists to identify esthetic risk factors, thereby improving implant treatment outcomes.

PMID:39892738 | DOI:10.1016/j.jdent.2025.105601

Neuroimage. 2025 Jan 30:121064. doi: 10.1016/j.neuroimage.2025.121064. Online ahead of print.

ABSTRACT

In this study, we present a comprehensive pipeline to train and compare a broad spectrum of machine learning and deep learning brain clocks, integrating diverse preprocessing strategies and correction terms. Our analysis also includes established methodologies which have shown success in prior UK Biobank-related studies. For our analysis we used T1-weighted MRI scans and processed de novo all images via FastSurfer, transforming them into a conformed space for deep learning and extracting image-derived phenotypes for our machine learning approaches. We rigorously evaluated these approaches both as robust age predictors for healthy individuals and as potential biomarkers for various neurodegenerative conditions, leveraging data from the UK Biobank, ADNI, and NACC datasets. To this end we designed a statistical framework to assess age prediction performance, the robustness of the prediction across cohort variability (database, machine type and ethnicity) and its potential as a biomarker for neurodegenerative conditions. Results demonstrate that highly accurate brain age models, typically utilising penalised linear machine learning models adjusted with Zhang's methodology, with mean absolute errors under 1 year in external validation, can be achieved while maintaining consistent prediction performance across different age brackets and subgroups (e.g., ethnicity and MRI machine/manufacturer). Additionally, these models show strong potential as biomarkers for neurodegenerative conditions, such as dementia, where brain age prediction achieved an AUROC of up to 0.90 in distinguishing healthy individuals from those with dementia.

PMID:39892529 | DOI:10.1016/j.neuroimage.2025.121064

Magn Reson Imaging. 2025 Jan 30:110342. doi: 10.1016/j.mri.2025.110342. Online ahead of print.

ABSTRACT

Brain tumor growth is associated with angiogenesis, wherein the density of newly developed blood vessels indicates tumor progression and correlates with the tumor grade. Quantitative dynamic contrast enhanced-magnetic resonance imaging (DCE-MRI) has shown potential in brain tumor grading and treatment response assessment. Segmentation of large-blood-vessels is crucial for automatic and accurate tumor grading using quantitative DCE-MRI. Traditional manual and semi-manual rule-based large-blood-vessel segmentation methods are time-intensive and prone to errors. This study proposes a novel deep learning-based technique for automatic large-blood-vessel segmentation using Swin UNETR architectures and comparing it with U-Net and Attention U-Net architectures. The study employed MRI data from 187 brain tumor patients, with training, validation, and testing datasets sourced from two centers, two vendors, and two field-strength magnetic resonance scanners. To test the generalizability of the developed model, testing was also carried out on different brain tumor types, including lymphoma and metastasis. Performance evaluation demonstrated that Swin UNETR outperformed other models in segmenting large-blood-vessel regions (achieving Dice scores of 0.979, and 0.973 on training and validation sets, respectively, with test set performance ranging from 0.835 to 0.982). Moreover, most quantitative parameters showed significant differences (p < 0.05) between with and without large-blood-vessel. After large-blood-vessel removal, using both ground truth and predicted masks, the values of parameters in non-vascular tumoral regions were statistically similar (p > 0.05). The proposed approach has potential applications in improving the accuracy of automatic grading of tumors as well as in treatment planning.

PMID:39892479 | DOI:10.1016/j.mri.2025.110342