Aesthetic Plast Surg. 2025 May 27. doi: 10.1007/s00266-025-04938-1. Online ahead of print.

NO ABSTRACT

PMID:40425882 | DOI:10.1007/s00266-025-04938-1

Nat Med. 2025 May 27. doi: 10.1038/s41591-025-03715-6. Online ahead of print.

ABSTRACT

There is limited evidence supporting the feasibility of using omics and functional technologies to inform treatment decisions. Here we present results from a cohort of 116 melanoma patients in the prospective, multicentric observational Tumor Profiler (TuPro) precision oncology project. Nine independent technologies, mostly at single-cell level, were used to analyze 126 patient samples, generating up to 500 Gb of data per sample (40,000 potential markers) within 4 weeks. Among established and experimental markers, the molecular tumor board selected 54 to inform its treatment recommendations. In 75% of cases, TuPro-based data were judged to be useful in informing recommendations. Patients received either standard of care (SOC) treatments or highly individualized, polybiomarker-driven treatments (beyond SOC). The objective response rate in difficult-to-treat palliative, beyond SOC patients (n = 37) was 38%, with a disease control rate of 54%. Progression-free survival of patients with TuPro-informed therapy decisions was 6.04 months, (95% confidence interval, 3.75-12.06) and 5.35 months (95% confidence interval, 2.89-12.06) in ≥third therapy lines. The proof-of-concept TuPro project demonstrated the feasibility and relevance of omics-based tumor profiling to support data-guided clinical decision-making. ClinicalTrials.gov identifier: NCT06463509 .

PMID:40425842 | DOI:10.1038/s41591-025-03715-6

Sci Rep. 2025 May 27;15(1):18556. doi: 10.1038/s41598-025-03010-x.

ABSTRACT

Alzheimer's disease (AD) is the most prevalent type of dementia. It is linked with a gradual decline in various brain functions, such as memory. Many research efforts are now directed toward non-invasive procedures for early diagnosis because early detection greatly benefits the patient care and treatment outcome. Additional to an accurate diagnosis and reduction of the rate of misdiagnosis; Computer-Aided Design (CAD) systems are built to give definitive diagnosis. This paper presents a novel CAD system to determine stages of AD. Initially, deep learning techniques are utilized to extract features from the AD brain MRIs. Then, the extracted features are reduced using a proposed feature reduction technique named Deep Downsized Kernel Partial Least Squares (DDKPLS). The proposed approach selects a reduced number of samples from the initial information matrix. The samples chosen give rise to a new data matrix further processed by KPLS to deal with the high dimensionality. The reduced feature space is finally classified using ELM. The implementation is named DDKPLS-ELM. Reference tests have been performed on the Kaggle MRI dataset, which exhibit the efficacy of the DDKPLS-based classifier; it achieves accuracy up to 95.4% and an F1 score of 95.1%.

PMID:40425715 | DOI:10.1038/s41598-025-03010-x

Sci Rep. 2025 May 27;15(1):18552. doi: 10.1038/s41598-025-02739-9.

ABSTRACT

Markerless motion capture (ML) systems, which utilize deep learning algorithms, have significantly expanded the applications of biomechanical analysis. Jump tests are now essential tools for athlete monitoring and injury prevention. However, the validity of kinematic and kinetic parameters derived from ML for lower limb joints requires further validation in populations engaged in high-intensity jumping sports. The purpose of this study was to compare lower limb kinematic and kinetic estimates between marker-based (MB) and ML motion capture systems during jumps. Fourteen male Division I movement collegiate athletes performed a minimum of three squat jumps (SJ), drop jumps (DJ), and countermovement jumps (CMJ) in a fixed sequence. The movements were synchronized using ten infrared cameras, six high-resolution cameras, and two force measurement platforms, all controlled by Vicon Nexus software. Motion data were collected, and the angles, moments, and power at the hip, knee, and ankle joints were calculated using Theia3D software. These results were then compared with those obtained from the Vicon system. Comparative analyses included Pearson correlation coefficients (r), root mean square differences (RMSD), extreme error values, and statistical parametric mapping (SPM).SPM analysis of the three movements in the sagittal plane revealed significant differences in hip joint angles, with joint angle RMSD ≤ 5.6°, hip joint moments RMSD ≤ 0.26 N·M/kg, and power RMSD ≤ 2.12 W/kg showing considerable variation, though not reaching statistical significance. ML systems demonstrate high measurement accuracy in estimating knee and ankle kinematics and kinetics in the sagittal plane during these conventional jump tests; however, the accuracy of hip joint kinematic measurements in the sagittal plane requires further validation.

PMID:40425708 | DOI:10.1038/s41598-025-02739-9

Thorax. 2025 May 27:thorax-2024-222060. doi: 10.1136/thorax-2024-222060. Online ahead of print.

ABSTRACT

BACKGROUND: Idiopathic pulmonary fibrosis (IPF) is a progressive disease of lung parenchymal scarring that is triggered by repeated microinjury to a vulnerable alveolar epithelium. It is increasingly recognised that cellular ageing, whether physiological or accelerated due to telomere dysfunction, renders the epithelium less able to cope with injury and triggers changes in epithelial behaviour that ultimately lead to the development of disease.

AIMS: This review aims to highlight how, with increasing age, the alveolar epithelium becomes vulnerable to exogenous insults. We discuss the downstream consequences of alveolar epithelial dysfunction on epithelial phenotype, alveolar repair and pro-pathogenic interactions with other alveolar niche-resident cell types which drive IPF pathogenesis.

NARRATIVE: We highlight how a wide array of cellular mechanisms that maintain cellular homeostasis become dysfunctional with ageing. Waning replicative capacity, genomic stability, mitochondrial function, proteostasis and metabolic function all contribute to a phenotype of vulnerability to 'second hits'. We discuss how in IPF the alveolar epithelium becomes dysfunctional, highlighting changes in repair capacity and fundamental cellular phenotype and how interactions between abnormal epithelium and other alveolar niche-resident cell types perpetuate disease.

CONCLUSIONS: The ageing epithelium is a vulnerable epithelium which, with the cumulative effects of environmental exposures, fundamentally changes its behaviour towards stalled differentiation, failed repair and profibrotic signalling. Further dissection of aberrant epithelial behaviour, and its impact on other alveolar cell types, will allow identification of novel therapeutic targets aimed at earlier pathogenic events.

PMID:40425299 | DOI:10.1136/thorax-2024-222060

Thorax. 2025 May 27:thorax-2025-223227. doi: 10.1136/thorax-2025-223227. Online ahead of print.

NO ABSTRACT

PMID:40425298 | DOI:10.1136/thorax-2025-223227

Toxicol Appl Pharmacol. 2025 May 25:117412. doi: 10.1016/j.taap.2025.117412. Online ahead of print.

ABSTRACT

PURPOSE: There is a need to generate new treatments against pulmonary diseases such as idiopathic fibrosis and asthma. N-acetylcysteine (NAC) has multiple clinical applications, but its unstable nature and route of administration limits its effectiveness. New pulmonary delivery strategies, such as liposomes made of lung surfactant lipids, could overcome NAC's limitations. This work aims to evaluate the efficacy of NAC combined with liposomes as a treatment for asthma and in preventing fibrotic development.

METHODS: Unilamellar vesicles were obtained through the dehydration-rehydration method followed by multiple membrane extrusion and characterized by Dynamic Light Scattering and Transmission electron microscopy. Lung fibrosis was induced by bleomycin administration, and liposomal formulation of NAC (LipoNAC) was evaluated as a preventive treatment. LipoNAC formulation was also evaluated in a therapeutic regimen for asthma using the classic ovalbumin model. For both models, the administration of the treatment was via the intranasal route.

RESULTS: NAC treatments (free NAC and LipoNAC) improved lung histopathology and decreased collagen deposition when tested in the lung fibrosis model. Only LipoNAC decreased serum levels of lactate dehydrogenase, myeloperoxidase activity in lung fluid and lung TGF-β. Although both treatments decreased Th2 cytokine and histopathological inflammation in the asthma model, only LipoNAC treatment significantly decreased mucus in asthmatic mice.

CONCLUSIONS: These results indicate that surfactant liposomal delivery of NAC potentiates its anti-inflammatory, mucolytic, and antioxidant activity, rendering it a promising therapy for respiratory diseases.

PMID:40425069 | DOI:10.1016/j.taap.2025.117412

Biochim Biophys Acta Mol Basis Dis. 2025 May 26;1871(7):167924. doi: 10.1016/j.bbadis.2025.167924. Online ahead of print.

NO ABSTRACT

PMID:40424856 | DOI:10.1016/j.bbadis.2025.167924

EMBO J. 2025 May 27. doi: 10.1038/s44318-025-00466-5. Online ahead of print.

ABSTRACT

The three-dimensional (3D) chromatin structure of Epstein-Barr virus (EBV) within host cells and the underlying mechanisms of chromatin interaction and gene regulation, particularly those involving EBV's noncoding RNAs (ncRNAs), have remained incompletely characterized. In this study, we employed state-of-the-art techniques of 3D genome mapping, including protein-associated chromatin interaction analysis with paired-end tag sequencing (ChIA-PET), RNA-associated chromatin interaction technique (RDD), and super-resolution microscopy, to delineate the spatial architecture of EBV in human lymphoblastoid cells. We systematically analyzed EBV-to-EBV (E-E), EBV-to-host (E-H), and host-to-host (H-H) interactions linked to host proteins and EBV RNAs. Our findings reveal that EBV utilizes host CCCTC-binding factor (CTCF) and RNA polymerase II (RNAPII) to form distinct chromatin contact domains (CCDs) and RNAPII-associated interaction domains (RAIDs). The anchors of these chromatin domains serve as platforms for extensive interactions with host chromatin, thus modulating host gene expression. Notably, EBV ncRNAs, especially Epstein-Barr-encoded RNAs (EBERs), target and interact with less accessible regions of host chromatin to repress a subset of genes via the inhibition of RNAPII-associated chromatin loops. This process involves the cofactor nucleolin (NCL) and its RNA recognition motifs, and depletion of either NCL or EBERs alters expression of genes crucial for host infection control, immune response, and cell cycle regulation. These findings unveil a sophisticated interplay between EBV and host chromatin.

PMID:40425856 | DOI:10.1038/s44318-025-00466-5

Nat Med. 2025 May 27. doi: 10.1038/s41591-025-03715-6. Online ahead of print.

ABSTRACT

There is limited evidence supporting the feasibility of using omics and functional technologies to inform treatment decisions. Here we present results from a cohort of 116 melanoma patients in the prospective, multicentric observational Tumor Profiler (TuPro) precision oncology project. Nine independent technologies, mostly at single-cell level, were used to analyze 126 patient samples, generating up to 500 Gb of data per sample (40,000 potential markers) within 4 weeks. Among established and experimental markers, the molecular tumor board selected 54 to inform its treatment recommendations. In 75% of cases, TuPro-based data were judged to be useful in informing recommendations. Patients received either standard of care (SOC) treatments or highly individualized, polybiomarker-driven treatments (beyond SOC). The objective response rate in difficult-to-treat palliative, beyond SOC patients (n = 37) was 38%, with a disease control rate of 54%. Progression-free survival of patients with TuPro-informed therapy decisions was 6.04 months, (95% confidence interval, 3.75-12.06) and 5.35 months (95% confidence interval, 2.89-12.06) in ≥third therapy lines. The proof-of-concept TuPro project demonstrated the feasibility and relevance of omics-based tumor profiling to support data-guided clinical decision-making. ClinicalTrials.gov identifier: NCT06463509 .

PMID:40425842 | DOI:10.1038/s41591-025-03715-6

Commun Biol. 2025 May 27;8(1):815. doi: 10.1038/s42003-025-08109-5.

ABSTRACT

Biological methods are a promising route for the environmentally-friendly production of rare earth elements (REE), which are essential for sustainable energy and defense technologies. In earlier work we identified the key genetic mechanisms contributing to the REE-bioleaching capability of Gluconobacter oxydans B58. Here we have targeted two of these mechanisms to generate a high-efficiency bioleaching strain of G. oxydans. Disruption of the phosphate-specific transport system through a clean deletion of pstS constitutively turns on the phosphate starvation response, yielding a much more acidic biolixiviant, and increasing bioleaching by up to 30%. Coupling knockout of pstS with the over-expression of the mgdh membrane-bound glucose dehydrogenase gene using the P112 promoter (strain G. oxydans ΔpstS, P112:mgdh) reduces biolixiviant pH by 0.39 units; increases REE-bioleaching by 53% at a pulp density of 10% and increases it by 73% at a pulp density of 1%.

PMID:40425722 | DOI:10.1038/s42003-025-08109-5

NPJ Sci Food. 2025 May 27;9(1):87. doi: 10.1038/s41538-025-00443-6.

ABSTRACT

Grape polyphenols (GPs) are rich in B-type proanthocyanidins, which promote metabolic resilience. Longitudinal metabolomic, metagenomic, and metaproteomic changes were measured in 27 healthy subjects supplemented with soy protein isolate (SPI, 40 g per day) for 5 days followed by GPs complexed to SPI (GP-SPI standardized to 5% GPs, 40 g per day) for 10 days. Fecal, urine, and/or fasting blood samples were collected before supplementation (day -5), after 5 days of SPI (day 0), and after 2, 4 and 10 days of GP-SPI. Most multi-omic changes observed after 2 and/or 4 days of GP-SPI intake were temporary, returning to pre-supplementation profiles by day 10. Shotgun metagenomics sequencing provided insights that could not be captured with 16S rRNA amplicon sequencing. Notably, 10 days of GP-SPI decreased fasting blood glucose and increased serum hyocholic acid (HCA), a glucoregulatory bile acid, which negatively correlated with one gut bacterial guild. In conclusion, GP-induced suppression of a bacterial guild may lead to higher HCA and lower fasting blood glucose.

PMID:40425565 | DOI:10.1038/s41538-025-00443-6

Nat Commun. 2025 May 27;16(1):4889. doi: 10.1038/s41467-025-59564-x.

ABSTRACT

Systematic inference of enzyme activity in human tumors is key to understanding cancer progression and resistance to therapy. However, standard protein or transcript abundances are blind to the activity status of the measured enzymes, regulated, for example, by active-site amino acid mutations or post-translational protein modifications. Current methods for activity-based proteome profiling (ABPP), which combine mass spectrometry (MS) with chemical probes, quantify the fraction of enzymes that are catalytically active. Here, we describe depletion-dependent ABPP (dd-ABPP) combined with automated SWATH/DIA-MS, which simultaneously determines three molecular layers of studied enzymes: i) catalytically active enzyme fractions, ii) enzyme and background protein abundances, and iii) context-dependent enzyme-protein interactions. We demonstrate the utility of the method in advanced lung adenocarcinoma (LUAD) by monitoring nearly 4000 protein groups and 200 serine hydrolases (SHs) in tumor and adjacent tissue sections routinely collected for patient histopathology. The activity profiles of 23 SHs and the abundance of 59 proteins associated with these enzymes retrospectively classified aggressive LUAD. The molecular signature revealed accelerated lipoprotein depalmitoylation via palmitoyl(protein)hydrolase activities, further confirmed by excess palmitate and its metabolites. The approach is universal and applicable to other enzyme families with available chemical probes, providing clinicians with a biochemical rationale for tumor sample classification.

PMID:40425563 | DOI:10.1038/s41467-025-59564-x

NPJ Syst Biol Appl. 2025 May 27;11(1):56. doi: 10.1038/s41540-025-00528-8.

ABSTRACT

Evolutionary cancer therapy (ECT) delays or forestalls the progression of metastatic cancer by adjusting treatment based on individual patient and disease characteristics. Clinical implementation of ECT can improve patient outcomes but faces technical and cultural challenges. To address those, we propose a systems approach incorporating systems modeling, problem structuring, and stakeholder engagement. This approach identifies and addresses barriers to implementation, ensuring the feasibility of ECT in clinical practice and enabling better metastatic cancer care.

PMID:40425536 | DOI:10.1038/s41540-025-00528-8

J R Soc Interface. 2025 May;22(226):20240852. doi: 10.1098/rsif.2024.0852. Epub 2025 May 28.

ABSTRACT

The long non-coding RNA metastasis-associated lung adenocarcinoma transcript 1 (lncRNA MALAT1) and microRNA-145-5p (miR-145) axis play a pivotal role in regulating drug resistance, apoptosis and senescence in non-small cell lung cancer (NSCLC). MALAT1 drives drug resistance by suppressing miR-145 and activating MUC1, thereby inhibiting ferroptosis; however, its precise role in regulating ferroptosis in NSCLC remains unclear. Therefore, we propose a computational modelling approach to unravel the impact of the MALAT1/miR-145 axis on ferroptosis and drug resistance, to identify potential therapeutic strategies that promote ferroptosis. Using Boolean logic and a stochastic updating scheme, we developed and validated a robust regulatory model that encompasses ferroptosis, apoptosis, senescence and drug resistance pathways. The model, based on extensive literature and validated through gain- and loss-of-function perturbations, demonstrated strong alignment with observed clinical data that were not included in its construction. Our analysis identified three previously unreported feedback loops, miR-145/Wip1/p53, miR-145/Myc/MALAT1 and miR-145/MUC1/BMI1, establishing miR-145 as a central regulator in NSCLC. Perturbations targeting MALAT1 and wild-type p53-induced phosphatase 1 (Wip1) revealed potential therapeutic opportunities, with miR-145 activation emerging as a promising strategy to induce ferroptosis and overcome drug resistance. These findings highlight the MALAT1/miR-145 axis as a transformative therapeutic target, presenting a computational foundation to advance NSCLC treatment strategies.

PMID:40425041 | DOI:10.1098/rsif.2024.0852

Sci Rep. 2025 May 27;15(1):18582. doi: 10.1038/s41598-025-03428-3.

ABSTRACT

A serious adverse reaction (SADR) may follow a vaccination against SARS-CoV-2 infection. We aimed to explore symptoms and reporting trends of SADRs to anti-SARS-CoV-2 vaccines based on the EudraVigilance database. This retrospective observational study analysed 250,966 suspected SADRs (with 62.8% reported in females), following the administration of 733,837,251 vaccine doses against SARS-CoV-2. Pfizer BioNTech (Comirnaty-Tozinameran), Moderna (Spikevax-Elastomeran), Janssen (Jcovden) and AstraZeneca (Vaxzevria) vaccines were analysed. The assessment included 897 types of SADRs across 12 categories. The most common clinical manifestations of SADRs to anti-SARS-CoV-2 vaccines vaccines encompassed neuropsychiatric (n = 121,877), cardiovascular (n = 78,167), as well as musculoskeletal and connective tissue disorders (n = 63,994). After summarising all SADRs, vaccination with Comirnaty was associated with the lowest risk of experiencing SADRs (754/million administered doses), followed by Spikevax (785/million doses), Jcovden (1,248/million doses) and Vaxzevria (2,301/million doses; p < 0.001). Regarding the vaccine administration timelines, the reporting of SADRs tends to be delayed and occurs over a longer time (p < 0.001). SADRs associated with anti-SARS-CoV-2 vaccines seem to be relatively rare. Compared to adenovirus-based vector vaccines (Jcovden, Vaxzevria), mRNA vaccines appear to offer improved safety profiles (Comirnaty, Spikevax). The risk of SADR to any SARS-CoV-2 vaccine seems to be outweighed by the benefits of active immunization against the virus.

PMID:40425703 | DOI:10.1038/s41598-025-03428-3

PLoS One. 2025 May 27;20(5):e0324293. doi: 10.1371/journal.pone.0324293. eCollection 2025.

ABSTRACT

Cotton is a major cash crop, and increasing its production is extremely important worldwide, especially in agriculture-led economies. The crop is susceptible to various diseases, leading to decreased yields. In recent years, advancements in deep learning methods have enabled researchers to develop automated methods for detecting diseases in cotton crops. Such automation not only assists farmers in mitigating the effects of the disease but also conserves resources in terms of labor and fertilizer costs. However, accurate classification of multiple diseases simultaneously in cotton remains challenging due to multiple factors, including class imbalance, variation in disease symptoms, and the need for real-time detection, as most existing datasets are acquired under controlled conditions. This research proposes a novel method for addressing these challenges and accurately classifying seven classes, including six diseases and a healthy class. We address the class imbalance issue through synthetic data generation using conventional methods like scaling, rotating, transforming, shearing, and zooming and propose a customized StyleGAN for synthetic data generation. After preprocessing, we combine features extracted from MobileNet and VGG16 to create a comprehensive feature vector, passed to three classifiers: Long Short Term Memory Units, Support Vector Machines, and Random Forest. We propose a StackNet-based ensemble classifier that takes the output probabilities of these three classifiers and predicts the class label among six diseases-Bacterial blight, Curl virus, Fusarium wilt, Alternaria, Cercospora, Greymildew-and a healthy class. We trained and tested our method on publicly available datasets, achieving an average accuracy of 97%. Our robust method outperforms state-of-the-art techniques to identify the six diseases and the healthy class.

PMID:40424461 | DOI:10.1371/journal.pone.0324293

Bioinformatics. 2025 May 27:btaf319. doi: 10.1093/bioinformatics/btaf319. Online ahead of print.

ABSTRACT

MOTIVATION: Drug side effects refer to harmful or adverse reactions that occur during drug use, unrelated to the therapeutic purpose. A core issue in drug side effect prediction is determining the frequency of these drug side effects in the population, which can guide patient medication use and drug development. Many computational methods have been developed to predict the frequency of drug side effects as an alternative to clinical trials. However, existing methods typically build regression models on five frequency classes of drug side effects and tend to overfit the training set, leading to boundary handling issues and the risk of overfitting.

RESULTS: To address this problem, we develop a multi-source similarity fusion-based model, named MSSF, for predicting five frequency classes of drug side effects. Compared to existing methods, our model utilizes the multi-source feature fusion module and the self-attention mechanism to explore the relationships between drugs and side effects deeply and employs Bayesian variational inference to more accurately predict the frequency classes of drug side effects. The experimental results indicate that MSSF consistently achieves superior performance compared to existing models across multiple evaluation settings, including cross-validation, cold-start experiments, and independent testing. The visual analysis and case studies further demonstrate MSSF's reliable feature extraction capability and promise in predicting the frequency classes of drug side effects.

AVAILABILITY: The source code of MSSF is available on GitHub (https://github.com/dingxlcse/MSSF.git) and archived on Zenodo (DOI: 10.5281/zenodo.15462041).

SUPPLEMENTARY INFORMATION: Additional files are available at Bioinformatics online.

PMID:40424358 | DOI:10.1093/bioinformatics/btaf319

PLoS One. 2025 May 27;20(5):e0322225. doi: 10.1371/journal.pone.0322225. eCollection 2025.

ABSTRACT

With the digital transformation of the financial industry, credit score prediction, as a key component of risk management, faces increasingly complex challenges. Traditional credit scoring methods often have difficulty in fully capturing the characteristics of large-scale, high-dimensional financial data, resulting in limited prediction performance. To address these issues, this paper proposes a credit score prediction model that combines CNNs and BiGRUs, and uses the GWO algorithm for hyperparameter tuning. CNN performs well in feature extraction and can effectively capture patterns in customer historical behaviors, while BiGRU is good at handling time dependencies, which further improves the prediction accuracy of the model. The GWO algorithm is introduced to further improve the overall performance of the model by optimizing key parameters. Experimental results show that the CNN-BiGRU-GWO model proposed in this paper performs well on multiple public credit score datasets, significantly improving the accuracy and efficiency of prediction. On the LendingClub loan dataset, the MAE of this model is 15.63, MAPE is 4.65%, RMSE is 3.34, and MSE is 12.01, which are 64.5%, 68.0%, 21.4%, and 52.5% lower than the traditional method plawiak of 44.07, 14.51%, 4.25, and 25.29, respectively. In addition, compared with traditional methods, this model also shows stronger advantages in adaptability and generalization ability. By integrating advanced technologies, this model not only provides an innovative technical solution for credit score prediction, but also provides valuable insights into the application of deep learning in the financial field, making up for the shortcomings of existing methods and demonstrating its potential for wide application in financial risk management.

PMID:40424348 | DOI:10.1371/journal.pone.0322225

PLoS One. 2025 May 27;20(5):e0322555. doi: 10.1371/journal.pone.0322555. eCollection 2025.

ABSTRACT

Human Action Recognition (HAR) has grown significantly because of its many uses, including real-time surveillance and human-computer interaction. Various variations in routine human actions make the recognition process of action more difficult. In this paper, we proposed a novel deep learning architecture known as Inverted Bottleneck Residual with Self-Attention (InBRwSA). The proposed architecture is based on two different modules. In the first module, 6-parallel inverted bottleneck residual blocks are designed, and each block is connected with a skip connection. These blocks aim to learn complex human actions in many convolutional layers. After that, the second module is designed based on the self-attention mechanism. The learned weights of the first module are passed to self-attention, extract the most essential features, and can easily discriminate complex human actions. The proposed architecture is trained on the selected datasets, whereas the hyperparameters are chosen using the particle swarm optimization (PSO) algorithm. The trained model is employed in the testing phase for the feature extraction from the self-attention layer and passed to the shallow wide neural network classifier for the final classification. The HMDB51 and UCF 101 are frequently used as action recognition standard datasets. These datasets are chosen to allow for meaningful comparison with earlier research. UCF101 dataset has a wide range of activity classes, and HMDB51 has varied real-world behaviors. These features test the generalizability and flexibility of the presented model. Moreover, these datasets define the evaluation scope within a particular domain and guarantee relevance to real-world circumstances. The proposed technique is tested on both datasets, and accuracies of 78.80% and 91.80% were achieved, respectively. The ablation study demonstrated that a margin of error value of 70.1338 ± 3.053 (±4.35%) and 82.7813 ± 2.852 (±3.45%) for the confidence level 95%,1.960σx̄ is obtained for HMDB51 and UCF datasets respectively. The training time for the highest accuracy for HDMB51 and UCF101 is 134.09 and 252.10 seconds, respectively. The proposed architecture is compared with several pre-trained deep models and state-of-the-art (SOTA) existing techniques. Based on the results, the proposed architecture outperformed existing techniques.

PMID:40424287 | DOI:10.1371/journal.pone.0322555