Cell Genom. 2025 Apr 17:100853. doi: 10.1016/j.xgen.2025.100853. Online ahead of print.

ABSTRACT

Mushroom-forming fungi (Agaricomycetes) are emerging as pivotal players in several fields of science and industry. Genomic data for Agaricomycetes are accumulating rapidly; however, this is not paralleled by improvements of gene annotations, which leave gene function notoriously poorly understood. We set out to improve our functional understanding of the model mushroom Coprinopsis cinerea by integrating a new, chromosome-level assembly, high-quality gene predictions, and functional information derived from broad gene-expression profiling data. The new annotation includes 5' and 3' untranslated regions (UTRs), polyadenylation sites (PASs), upstream open reading frames (uORFs), splicing isoforms, and microexons, as well as core gene sets corresponding to carbon starvation, light response, and hyphal differentiation. As a result, the genome of C. cinerea has now become the most comprehensively annotated genome among mushroom-forming fungi, which will contribute to multiple rapidly expanding fields, including research on their life history, light and stress responses, as well as multicellular development.

PMID:40262612 | DOI:10.1016/j.xgen.2025.100853

Curr Biol. 2025 Apr 21;35(8):R298-R300. doi: 10.1016/j.cub.2025.03.021.

ABSTRACT

Duckweeds are among the smallest and fastest-growing flowering plants. In a new study that combines experimental data with phylogenomic comparisons across the clade, the authors explore how changes in gene content, epigenetic pathways, and their interplay shaped the body plan, aquatic lifestyle, and clonal growth habit of this plant family.

PMID:40262538 | DOI:10.1016/j.cub.2025.03.021

Cognition. 2025 Apr 21;261:106144. doi: 10.1016/j.cognition.2025.106144. Online ahead of print.

ABSTRACT

The ability to learn new environments is a foundational human skill, yet we know little about how exploration behaviors shape spatial learning. Here, we investigated the relationships between exploration behaviors and spatial memory in healthy young adults, and further related performance to other measures of individual differences. In the present study, 100 healthy young adults (ages 18-37) freely explored a maze in a virtual desktop environment to learn the locations of 9 objects. Participants then navigated from one object to another without feedback, and their accuracy and path efficiency were determined. Interestingly, participant accuracy ranged from near 0 % to 100 %. Correlations and principal component regression revealed that evenness of exploration (i.e., visiting all locations with a similar frequency) and how quickly all objects were found during exploration were related to performance. Indeed, differences in performance become apparent by the time participants found the 6th object (within the first 50 moves), emphasizing the importance of exploration quality over exploration quantity. Perspective taking ability and video game experience were also related to performance. Critically, we found no correlations between performance on matched pairs of active-passive exploration paths, suggesting that experiencing a "good" exploration path does not lead to better performance; instead, the path is more likely a reflection of the navigator's ability. Sex differences were observed, however, a serial mediation analysis revealed that even exploration had a greater explanatory effect on those sex differences compared to video game experience. Our results indicate that exploration behaviors predict navigational performance and highlight the importance of moment-to-moment behaviors exhibited during exploration and learning.

PMID:40262422 | DOI:10.1016/j.cognition.2025.106144

J Hazard Mater. 2025 Apr 17;493:138330. doi: 10.1016/j.jhazmat.2025.138330. Online ahead of print.

ABSTRACT

Polycarbonate (PC), a widely used thermoplastic, poses significant environmental challenges due to its persistence and the release of bisphenol A (BPA), a known xenoestrogen. Here, we report the isolation of Bacillus subtilis JNU01 (BsJNU01), capable of utilizing PC as its sole carbon source. Through transcriptomic analysis, we identified metallophosphoesterase from BsJNU01 (BsMPPE), the first reported metallophosphoesterase capable of degrading polycarbonate by catalyzing the hydrolysis of carbonate ester bonds. This enzyme operates under mild aqueous conditions (30 °C, pH 7), releasing 30 μmol of BPA as a monomer and demonstrating effective PC degradation under environmentally friendly conditions. PC biodegradation was confirmed by Fourier transform infrared spectroscopy (FT-IR), nuclear magnetic resonance (NMR), and gas chromatography-mass spectrometry (GC-MS). Furthermore, surface and mechanical analyses revealed significant degradation and structural changes in PC films following BsMPPE treatment, with toughness showing a 40-70 % decrease compared to untreated PC films. This study represents a breakthrough in microbial plastic degradation, establishing a sustainable biocatalytic platform for PC recycling and upcycling technologies.

PMID:40262317 | DOI:10.1016/j.jhazmat.2025.138330

Emerg Microbes Infect. 2025 Apr 22:2497302. doi: 10.1080/22221751.2025.2497302. Online ahead of print.

ABSTRACT

ABSTRACTChronic obstructive pulmonary disease (COPD) and aging both increase the risk of tuberculosis (TB), an important infectious disease in human. Exploring the burden and predictors of latent tuberculosis infection (LTBI) and treatment outcomes for older individuals with COPD is essential to guide LTBI intervention policy. We enrolled patients aged over 60 years with COPD between January 2021 and June 2023 for LTBI screening using interferon-gamma release assay (IGRA). LTBI treatment options included all WHO-recommended regimens. The final regimen was selected through shared decision-making between patients and their COPD physicians, leveraging the long-standing rapport being established. We investigated the prevalence of LTBI in this population, identified risk factors using logistic regression analysis, and evaluated treatment outcomes. A total of 810 COPD patients (mean: 72.8-years) underwent LTBI screening, with an IGRA-positive rate of 23.8%. IGRA positivity was correlated with smoking pack-years (adjusted odds ratio [aOR]: 1.02, p < 0.001), current smoking status (aOR 1.40, p = 0.030), COPD duration (aOR 1.10, p = 0.03), inhaled corticosteroid use (aOR 3.06, p < 0.001), and a cumulative equivalent dose of prednisolone exceeding 210 mg over 2 years (aOR 3.13, p < 0.001). Treatment was initiated in 150 patients (77.7%), predominantly with weekly rifapentine plus isoniazid (3HP) (60.7%). The overall completion rate was 82.0%, with adverse reactions being the primary reason for discontinuation. Our findings support that the LTBI intervention is recommended for older patients with COPD, especially those at higher risk, as nearly 25% of them have tuberculosis infection. The high treatment completion rate highlights the safety and feasibility of the WHO-recommended regimens.

PMID:40262275 | DOI:10.1080/22221751.2025.2497302

Mol Biol Rep. 2025 Apr 22;52(1):410. doi: 10.1007/s11033-025-10524-0.

ABSTRACT

Drug repurposing is the process of using currently approved drugs for a novel treatment or medical condition for which it was not previously indicated. Despite promising preclinical and clinical results, most of the newly designed senotherapeutic agents synthesized have limited clinical utility due to individual and organ-specific variations in aging phenotype and adverse side effects. All these limitations indicate that further clinical research is required to determine the effectiveness of repurposed senotherapeutic drug interventions, such as metformin, for age-related diseases. Metformin exerts diverse senotherapeutic effects on various aging tissues at different metabolic conditions. Although not exhibiting senolytic properties, metformin has effectively suppressed cellular senescence and senescence-associated secretory phenotype (SASP) in age-related diseases (ARDs). Targeting specific SASP-related signaling pathways with metformin may offer new therapeutic benefits to alleviate the detrimental effects of senescent cells accumulated in most common ARDs in the elderly. Metformin was also the first drug evaluated for its senescence-targeting effects in a large clinical trial named "Targeting Aging with Metformin (TAME)". In this review, we critically evaluate the literature to highlight senotherapeutic mechanisms in which metformin can be therapeutically repurposed for the prevention and treatment of ARDs.

PMID:40261556 | DOI:10.1007/s11033-025-10524-0

Clin Transl Sci. 2025 Apr;18(4):e70215. doi: 10.1111/cts.70215.

ABSTRACT

Clinical pharmacologists face unique challenges when developing drugs for rare diseases. These conditions are characterized by small patient populations, diverse disease progression patterns, and a limited understanding of underlying pathophysiology. This tutorial serves as a comprehensive guide, offering practical insights and strategies to navigate its complexities. In this tutorial, we outline global regulatory incentives and resources available to support rare disease research, describe some considerations for designing a clinical development plan for rare diseases, and we highlight the role of biomarkers, real-world data, and modeling and simulations to navigate rare disease challenges. By leveraging these tools and understanding regulatory pathways, clinical pharmacologists can significantly contribute to advancing therapeutic options for rare diseases.

PMID:40261641 | DOI:10.1111/cts.70215

Psychiatry Res. 2025 Apr 19;348:116506. doi: 10.1016/j.psychres.2025.116506. Online ahead of print.

ABSTRACT

Non-specific response to treatment (NSRT) is the primary contributor to the failure of randomized clinical trials in major depressive disorder (MDD). The objective of this study is to develop artificial neural network (ANN) models to predict the individual probability for NSRT. Pre-randomization data from a failed antidepressant trial were considered as potential predictors of the NSRT probability (prob-NSRT) using the response endpoint in subjects randomized to placebo. The inverse of the individual prob-NSRT (NSRT propensity score) was used as a weight in the mixed-effects model applied to assess treatment effect (TE). The comparison of the results obtained with and without the NSRT propensity score indicated that the weighted analyses provided an estimate of TE significantly larger than the conventional analyses. The propensity score weighted (PSW) analysis, adjusting for inter-individual variability in prob-NSRT, enhanced signal detection of TE. These findings support the potential role of PSW methodology for analyzing RCTs and determining TE. However, external validation of these ANN models in at least one independent trial is needed before advocating regulatory or broader clinical use.

PMID:40262198 | DOI:10.1016/j.psychres.2025.116506

Clin Pharmacol Ther. 2025 Apr 22. doi: 10.1002/cpt.3680. Online ahead of print.

ABSTRACT

The polymorphic CYP2D6 enzyme plays a pivotal role in the metabolism of approximately 25% of clinically prescribed drugs. However, the impact of specific genetic variants on the interindividual variability in CYP2D6-mediated drug metabolism remains insufficiently quantified. This translational study sought to address this gap by analyzing the genotypes and phenotypes of patients in two large clinical cohorts, focusing on the metabolism of the CYP2D6 substrates risperidone and desmethyltamoxifen. The analysis incorporated novel polymorphic haplotypes and substrate-specific differences among the CYP2D6.1, CYP2D6.2, and CYP2D6.35 enzyme variants. The study revealed that CYP2D6.2 and CYP2D6.35 exhibit reduced metabolic capacity for these substrates, both in vivo and in an in vitro expression model. This was evidenced by decreased catalytic turnover (Kcat), decreased substrate affinity, and altered substrate docking. Furthermore, novel polymorphic haplotypes on the CYP2D6*1, CYP2D6*2, and CYP2D6*35 backgrounds were identified, each associated with a 30-40% increase in CYP2D6 activity. Incorporating these findings into prediction equations significantly improved the genetic prediction accuracy (R2) for CYP2D6-mediated metabolism of desmethyltamoxifen from 59% to 71% and risperidone, also metabolized by CYP3A4, from 42% to 46%. These results highlight the importance of accounting for drug-specific interactions with enzyme variants and integrating distinct polymorphic haplotypes into CYP pharmacogenomic models and guidelines for better translation into clinical practice.

PMID:40261922 | DOI:10.1002/cpt.3680

Appl Microbiol Biotechnol. 2025 Apr 22;109(1):97. doi: 10.1007/s00253-025-13481-7.

ABSTRACT

Previous studies have established a correlation between the microbiome-derived metabolite trimethylamine N-oxide (TMAO) and decreased renal function, but with great heterogeneity. Moreover, population-based evidence remains scarce, particularly in Chinese populations. We designed a meta-analysis and a population-based cross-sectional study in China to examine the associations between TMAO and chronic kidney disease (CKD). In meta-analysis, among 2125 pooled subjects with 1240 controls and 885 CKD patients, a significant association was observed between TMAO and CKD, with a standardized mean difference of - 0.93 (95% confidence interval: - 1.11, - 0.75). Meta-regression analysis identified gender, age, and body mass index (BMI) as significant heterogeneity factors. In our population-based study of 5584 subjects with an estimated glomerular filtration rate (eGFR) ≥ 60 mL/min/1.73 m2 from Sijing community, 100 developed CKD in 2 years. We matched 195 controls by age and gender from the 5484 non-CKD subjects. Male subjects and alcohol consumers exhibited a lower risk of CKD with adjusted odds ratio (OR) of 0.471 (P < 0.05) and 0.320 (P < 0.05), respectively. When comparing subjects in the lowest tertile of TMAO, adjusted OR reached to 1.243 (P > 0.05) for those in the middle and 2.123 (P < 0.05) in the highest tertile (P for trend < 0.05). TMAO demonstrated a moderate capacity to distinguish CKD from non-CKD subjects (AUC = 0.614, P < 0.01). Our findings indicate TMAO is significantly associated with the risk of CKD, and suggest age, gender, and BMI may confound the relationship between TMAO and CKD. KEY POINTS: • Subjects with elevated TMAO levels have an increased risk of CKD. • TMAO demonstrates a moderate capacity to distinguish CKD from non-CKD cases. • Age, gender and BMI may confound the relationship between TMAO and CKD.

PMID:40261397 | DOI:10.1007/s00253-025-13481-7

Proc Natl Acad Sci U S A. 2025 Apr 29;122(17):e2418407122. doi: 10.1073/pnas.2418407122. Epub 2025 Apr 22.

ABSTRACT

Cystic fibrosis (CF) is a lethal genetic disorder caused by variants in CF transmembrane conductance regulator (CFTR). Many variants are treatable with correctors, which enhance the folding and trafficking of CFTR. However, approximately 3% of persons with CF harbor poorly responsive variants. Here, we used affinity purification mass spectrometry proteomics to profile the protein homeostasis (proteostasis) changes of CFTR variants during correction to assess modulated interactions with protein folding and maturation pathways. Responsive variant interactions converged on similar proteostasis pathways during correction. In contrast, poorly responsive variants subtly diverged, revealing a partial restoration of protein quality control surveillance and partial correction. Computational structural modeling showed that corrector VX-445 failed to confer enough NBD1 stability to poor responders. NBD1 secondary stabilizing mutations rescued poorly responsive variants, revealing structural vulnerabilities in NBD1 required for treating poor responders. Our study provides a framework for discerning the underlying protein quality control and structural defects of CFTR variants not reached with existing drugs to expand therapeutics to all susceptible CFTR variants.

PMID:40261935 | DOI:10.1073/pnas.2418407122

PLoS Pathog. 2025 Apr 22;21(4):e1013067. doi: 10.1371/journal.ppat.1013067. eCollection 2025 Apr.

NO ABSTRACT

PMID:40261841 | DOI:10.1371/journal.ppat.1013067

JCI Insight. 2025 Apr 22:e188146. doi: 10.1172/jci.insight.188146. Online ahead of print.

ABSTRACT

Pf bacteriophages, lysogenic viruses that infect Pseudomonas aeruginosa (Pa), are implicated in the pathogenesis of chronic Pa infections; phage-infected (Pf+) strains are known to predominate in people with cystic fibrosis (pwCF) who are older and have more severe disease. However, the transmission patterns of Pf underlying the progressive dominance of Pf+ strains are unclear. In particular, it is unknown whether phage transmission commonly occurs horizontally between bacteria via viral particles within the airway or if Pf+ bacteria are mostly acquired via de novo Pseudomonas infections. Here, we studied Pa genomic sequences from 3 patient cohorts totaling 662 clinical isolates from 105 pwCF. We identified Pf+ isolates and analyzed transmission patterns of Pf within patients between genetically similar groups of bacteria called "clone types". We found that Pf was predominantly passed down vertically within Pa clone types and rarely via horizontal transfer between clone types within the airway. Conversely, we found extensive evidence of Pa de novo infection by a new, genetically distinct Pf+ Pa. Finally, we observed that clinical isolates showed reduced activity of the type IV pilus and reduced susceptibility to Pf in vitro. These results cast new light on the transmission of virulence-associated phages in the clinical setting.

PMID:40261708 | DOI:10.1172/jci.insight.188146

JCI Insight. 2025 Apr 22:e187951. doi: 10.1172/jci.insight.187951. Online ahead of print.

ABSTRACT

Pharmacological rescue of F508del-CFTR by the triple combination CFTR modulator therapy elexacaftor/tezacaftor/ivacaftor (ETI) leads to unprecedented clinical benefits in patients with cystic fibrosis (CF), however, previous studies in CF primary human airway epithelial cultures demonstrated that chronic treatment with the potentiator ivacaftor can render the F508del protein unstable thus limiting restoration of CFTR chloride channel function. However, quantitative studies of this unwanted effect of ivacaftor on F508del channel function including dependency on cell culture conditions remain limited and the impact of chronic ivacaftor exposure on restoration of mucociliary clearance that is impaired in patients with CF has not been studied. In patient-derived primary nasal epithelial cultures, we found that different culture conditions (UNC-ALI medium vs. PneumaCult medium) have profound effects on ETI-mediated restoration of F508del-CFTR function. Chronic treatment with ivacaftor as part of ETI triple therapy limited the rescue of F508del-CFTR chloride channel function when CF nasal epithelial cultures were grown in UNC-ALI medium, but not in PneumaCult medium. In PneumaCult medium, both chronic and acute addition of ivacaftor as part of ETI treatment led to constitutive CFTR-mediated chloride secretion in the absence of exogenous cAMP-dependent stimulation. This constitutive CFTR-mediated chloride secretion was essential to improve viscoelastic properties of the mucus layer and to restore mucociliary transport on CF nasal epithelial cultures. Furthermore, nasal potential difference measurements in patients with CF showed that ETI restored constitutive F508del-CFTR activity in vivo. These results demonstrate that ivacaftor as a component of ETI therapy is essential to restore mucociliary clearance and suggest that this effect is facilitated by its constitutive activation of F508del channels following their folding-correction in patients with CF.

PMID:40261705 | DOI:10.1172/jci.insight.187951

Physiother Theory Pract. 2025 Apr 22:1-8. doi: 10.1080/09593985.2025.2494114. Online ahead of print.

ABSTRACT

OBJECTIVE: To characterize physiological responses to a 1-minute sit-to-stand test (STS) and assess correlations with cardiopulmonary exercise test (CPET) variables, nutritional status, pulmonary function, and quadriceps muscle strength in cystic fibrosis (CF) patients.

METHODS: Subjects aged 6-18 years with a genetic diagnosis of CF were enrolled in this cross-sectional study. After collecting demographic, anthropometric, and clinical data the following tests were performed: pulmonary function (spirometry), aerobic fitness (CPET), STS, and isometric quadriceps muscle strength (hand-held dynamometry). Data collection was performed on the same day.

RESULTS: The study sample comprised 17 children (9.8 ± 1.6 years) and adolescents (13.7 ± 1.5 years) with a mean forced expiratory volume in one second (FEV1) of - 0.80 ± 1.61 (z-score). In the CPET, peak exercise oxygen consumption (VO2peak) was 35.1 ± 4.2 mL.kg-1.min-1, while in the STS mean number of repetitions was 32.5 ± 6.2 and total work (repetitions × body mass) was 1326.9 ± 379.6. At peak exercise, CPET elicited higher heart rate (p = .001) and subjective sensation of dyspnea (p = .001) compared to STS, though no significant differences were observed in peripheral oxygen saturation. Moderate and significant correlations were identified between total workload (CPET) and repetitions adjusted for body weight (r = 0.684; p = .002) and between STS repetitions and muscle strength corrected for body weight (r = 0.531; p = .034). No significant correlations were found with nutritional status (BMI), pulmonary function (FEV1), or other aerobic fitness variables (VO2 at ventilatory threshold or VO2peak).

CONCLUSION: In children and adolescents with CF, compared to CPET, the STS test elicits a submaximal cardiorespiratory response that is mostly dependent on quadriceps muscle strength.

PMID:40260956 | DOI:10.1080/09593985.2025.2494114

J Xray Sci Technol. 2025 Apr 22:8953996251325092. doi: 10.1177/08953996251325092. Online ahead of print.

ABSTRACT

BackgroundPathological images play a crucial role in the diagnosis of critically ill cancer patients. Since cancer patients often seek medical assistance when their condition is severe, doctors face the urgent challenge of completing accurate diagnoses and developing surgical plans within a limited timeframe. The complexity and diversity of pathological images require a significant investment of time from specialized physicians for processing and analysis, which can lead to missing the optimal treatment window.PurposeCurrent medical decision support systems are challenged by the high computational complexity of deep learning models, which demand extensive data training, making it difficult to meet the real-time needs of emergency diagnostics.MethodThis study addresses the issue of emergency diagnosis for malignant bone tumors such as osteosarcoma by proposing a Lightened Boundary-enhanced Digital Pathological Image Recognition Strategy (LB-DPRS). This strategy optimizes the self-attention mechanism of the Transformer model and innovatively implements a boundary segmentation enhancement strategy, thereby improving the recognition accuracy of tissue backgrounds and nuclear boundaries. Additionally, this research introduces row-column attention methods to sparsify the attention matrix, reducing the computational burden of the model and enhancing recognition speed. Furthermore, the proposed complementary attention mechanism further assists convolutional layers in fully extracting detailed features from pathological images.ResultsThe DSC value of LB-DPRS strategy reached 0.862, the IOU value reached 0.749, and the params was only 10.97 M.ConclusionExperimental results demonstrate that the LB-DPRS strategy significantly improves computational efficiency while maintaining prediction accuracy and enhancing model interpretability, providing powerful and efficient support for the emergency diagnosis of malignant bone tumors such as osteosarcoma.

PMID:40262109 | DOI:10.1177/08953996251325092

J Chem Inf Model. 2025 Apr 22. doi: 10.1021/acs.jcim.5c00296. Online ahead of print.

ABSTRACT

In chemical reaction network theory, ordinary differential equations are used to model the temporal change of chemical species concentration. As the functional form of these ordinary differential equation systems is derived from an empirical model of the reaction network, it may be incomplete. Our approach aims to elucidate these hidden insights in the reaction network by combining dynamic modeling with deep learning in the form of neural ordinary differential equations. Our contributions not only help to identify the shortcomings of existing empirical models but also assist the design of future reaction networks.

PMID:40262040 | DOI:10.1021/acs.jcim.5c00296

Arch Insect Biochem Physiol. 2025 Aug;118(4):e70060. doi: 10.1002/arch.70060.

ABSTRACT

Goji berry is an important economic crop, yet pest infestations pose a significant threat to its yield and quality. Traditional pest identification mainly relies on manual inspection by experts with specialized knowledge, which is subjective, time-consuming, and labor-intensive. To address these issues, this experiment proposes an improved convolutional neural network (CNN) for accurate identification of 17 types of goji berry pests. Firstly, the original data set is augmented using a multi-graph-occlusion data augmentation method. Subsequently, the augmented data set is imported into the improved CNN for training. Based on the original ResNet18 model, a new CNN, named GojiNet, is constructed by embedding multi-attention fusion modules at appropriate locations. Experimental results demonstrate that GojiNet achieves an average recognition accuracy of 95.35%, representing a 2.60% improvement over the ResNet18 network. Notably, compared to the original network, the training time of this model increases only slightly, while its size is reduced, and the recognition accuracy is enhanced. The experiment verifies the performance of the GojiNet model through a series of evaluation indicators. This study confirms the tremendous potential and application prospects of deep learning in pest identification, providing a referential solution for intelligent and precise pest identification.

PMID:40262026 | DOI:10.1002/arch.70060

PLoS One. 2025 Apr 22;20(4):e0321849. doi: 10.1371/journal.pone.0321849. eCollection 2025.

ABSTRACT

Metallized Ceramic Ring is a novel electronic apparatus widely applied in communication, new energy, aerospace and other fields. Due to its complicated technique, there would be inevitably various defects on its surface; among which, the tiny pinhole defects with complex texture are the most difficult to detect, and there is no reliable method of automatic detection. This Paper proposes a method of detecting micro-pinhole defects on surface of metallized ceramic ring combining Improved Detection Transformer (DETR) Network with morphological operations, utilizing two modules, namely, deep learning-based and morphology-based pinhole defect detection to detect the pinholes, and finally combining the detection results of such two modules, so as to obtain a more accurate result. In order to improve the detection performance of DETR Network in aforesaid module of deep learning, EfficientNet-B2 is used to improve ResNet-50 of standard DETR network, the parameter-free attention mechanism (SimAM) 3-D weight attention mechanism is used to improve Sequeeze-and-Excitation (SE) attention mechanism in EfficientNet-B2 network, and linear combination loss function of Smooth L1 and Complete Intersection over Union (CIoU) is used to improve regressive loss function of training network. The experiment indicates that the recall and the precision of the proposed method are 83.5% and 86.0% respectively, much better than current mainstream methods of micro defect detection, meeting requirements of detection at industrial site.

PMID:40261923 | DOI:10.1371/journal.pone.0321849

PLoS One. 2025 Apr 22;20(4):e0321697. doi: 10.1371/journal.pone.0321697. eCollection 2025.

ABSTRACT

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder leading to an inability to socially communicate and in extreme cases individuals are completely dependent on caregivers. ASD detection at early ages is crucial as early detection can reduce the effect on social impairment. Deep learning models have shown capability to detect ASD earlier compared to traditional detection methods used by clinics and experts. Ensemble models, renowned for their ability to enhance predictive performance by combining multiple models, have emerged as a powerful tool in machine learning. This study harnesses the strength of ensemble learning to address the critical challenge of ASD diagnosis. This study proposed a deep ensemble model leveraging the strengths of VGG16 and Xception net trained on Facial Images for ASD detection overcoming limitations in existing datasets through extensive preprocessing. Proposed model preprocessed the training dataset of facial images by converting side posed images into frontal face images, using Histogram Equalization (HE) to enhance colors, data augmentation techniques application, and using the Hue Saturation Value (HSV) color model. By integrating the feature extraction strengths of VGG16 and Xception with fully connected layers, our model has achieved a notable 97% accuracy on the Kaggle ASD Face Image Dataset. This approach supports early detection of ASD and aligns with Sustainable Development Goal 3, which focuses on improving health and well-being.

PMID:40261913 | DOI:10.1371/journal.pone.0321697