Front Oncol. 2025 Jan 28;15:1528654. doi: 10.3389/fonc.2025.1528654. eCollection 2025.

ABSTRACT

OBJECTIVES: To develop a novel automatic delineation model, the Multi-Scale Channel Attention U-Net (MCAU-Net) model, for gallbladder segmentation on CT images of patients with liver cancer.

METHODS: We retrospectively collected the CT images from 120 patients with liver cancer, based on which ground truth was manually delineated by physicians. The images and ground truth constitute a dataset, which was proportionally divided into a training set (54%), a validation set (6%), and a test set (40%). Data augmentation was performed on the training set. Our proposed MCAU-Net model was employed for gallbladder segmentation and its performance was evaluated using Dice Similarity Coefficient (DSC), Jaccard Similarity Coefficient (JSC), Positive Predictive Value (PPV), Sensitivity (SE), Hausdorff Distance (HD), Relative Volume Difference (RVD), and Volumetric Overlap Error (VOE) metrics.

RESULTS: On the test set, MCAU-Net achieved DSC, JSC, PPV, SE, HD, RVD, and VOE values of 0.85 ± 0.22, 0.79 ± 0.23, 0.92 ± 0.14, 0.84 ± 0.23, 2.75 ± 0.98, 0.18 ± 0.48, and 0.22 ± 0.42, respectively. Compared to the control models, U-Net, SEU-Net and TransUNet, the MCAU-Net improved DSC 0.06, 0.04 and 0.06, JSC by 0.09, 0.06 and 0.09, PPV by 0.08, 0.08 and 0.05, SE by 0.05,0.05 and 0.07, and reduced HD by 0.45, 0.28 and 0.41, RVD by 0.07, 0.03 and 0.07, VOE by 0.04, 0.02 and 0.08 respectively. Qualitative results revealed that MCAU-Net produced smoother and more accurate boundaries, closer to the expert delineation, with less over-segmentation and under-segmentation and improved robustness.

CONCLUSIONS: The MCAU-Net model significantly improves gallbladder segmentation on CT images. It satisfies clinical requirements and enhances the efficiency of physicians, particularly in segmenting complex anatomical structures.

PMID:39935828 | PMC:PMC11810919 | DOI:10.3389/fonc.2025.1528654

Front Med (Lausanne). 2025 Jan 28;12:1492709. doi: 10.3389/fmed.2025.1492709. eCollection 2025.

ABSTRACT

OBJECTIVE: To systematically map the knowledge landscape and development trends in artificial intelligence (AI) applications for antimicrobial resistance (AMR) research through bibliometric analysis, providing evidence-based insights to guide future research directions and inform strategic decision-making in this dynamic field.

METHODS: A comprehensive bibliometric analysis was performed using the Web of Science Core Collection database for publications from 2014 to 2024. The analysis integrated multiple bibliometric approaches: VOSviewer for visualization of collaboration networks and research clusters, CiteSpace for temporal evolution analysis, and quantitative analysis of publication metrics. Key bibliometric indicators including co-authorship patterns, keyword co-occurrence, and citation impact were analyzed to delineate research evolution and collaboration patterns in this domain.

RESULTS: A collection of 2,408 publications was analyzed, demonstrating significant annual growth with publications increasing from 4 in 2014 to 549 in 2023 (22.7% of total output). The United States (707), China (581), and India (233) were the leading contributors in international collaborations. The Chinese Academy of Sciences (53), Harvard Medical School (43), and University of California San Diego (26) were identified as top contributing institutions. Citation analysis highlighted two major breakthroughs: AlphaFold's protein structure prediction (6,811 citations) and deep learning approaches to antibiotic discovery (4,784 citations). Keyword analysis identified six enduring research clusters from 2014 to 2024: sepsis, artificial neural networks, antimicrobial resistance, antimicrobial peptides, drug repurposing, and molecular docking, demonstrating the sustained integration of AI in antimicrobial therapy development. Recent trends show increasing application of AI technologies in traditional approaches, particularly in MALDI-TOF MS for pathogen identification and graph neural networks for large-scale molecular screening.

CONCLUSION: This bibliometric analysis shows the importance of artificial intelligence in enhancing the progress in the discovery of antimicrobial drugs especially toward the fight against AMR. From enhancing the fast, efficient and predictive performance of drug discovery methods, current AI capabilities have revealed observable potential to be proactive in combating the ever-growing challenge of AMR worldwide. This study serves not only an identification of current trends, but also, and especially, offers a strategic approach to further investigations.

PMID:39935800 | PMC:PMC11810743 | DOI:10.3389/fmed.2025.1492709

Front Cell Dev Biol. 2025 Jan 28;12:1517240. doi: 10.3389/fcell.2024.1517240. eCollection 2024.

ABSTRACT

PURPOSE: Using deep learning model to observe the blinking characteristics and evaluate the changes and their correlation with tear film characteristics in children with long-term use of orthokeratology (ortho-K).

METHODS: 31 children (58 eyes) who had used ortho-K for more than 1 year and 31 age and gender-matched controls were selected for follow-up in our ophthalmology clinic from 2021/09 to 2023/10 in this retrospective case-control study. Both groups underwent comprehensive ophthalmological examinations, including Ocular Surface Disease Index (OSDI) scoring, Keratograph 5M, and LipiView. A deep learning system based on U-Net and Swim-Transformer was proposed for the observation of blinking characteristics. The frequency of incomplete blinks (IB), complete blinks (CB) and incomplete blinking rate (IBR) within 20 s, as well as the duration of the closing, closed, and opening phases in the blink wave were calculated by our deep learning system. Relative IPH% was proposed and defined as the ratio of the mean of IPH% within 20 s to the maximum value of IPH% to indicate the extent of incomplete blinking. Furthermore, the accuracy, precision, sensitivity, specificity, F1 score of the overall U-Net-Swin-Transformer model, and its consistency with built-in algorithm were evaluated as well. Independent t-test and Mann-Whitney test was used to analyze the blinking patterns and tear film characteristics between the long-term ortho-K wearer group and the control group. Spearman's rank correlation was used to analyze the relationship between blinking patterns and tear film stability.

RESULTS: Our deep learning system demonstrated high performance (accuracy = 98.13%, precision = 96.46%, sensitivity = 98.10%, specificity = 98.10%, F1 score = 0.9727) in the observation of blinking patterns. The OSDI scores, conjunctival redness, lipid layer thickness (LLT), and tear meniscus height did not change significantly between two groups. Notably, the ortho-K group exhibited shorter first (11.75 ± 7.42 s vs. 14.87 ± 7.93 s, p = 0.030) and average non-invasive tear break-up times (NIBUT) (13.67 ± 7.0 s vs. 16.60 ± 7.24 s, p = 0.029) compared to the control group. They demonstrated a higher IB (4.26 ± 2.98 vs. 2.36 ± 2.55, p < 0.001), IBR (0.81 ± 0.28 vs. 0.46 ± 0.39, p < 0.001), relative IPH% (0.3229 ± 0.1539 vs. 0.2233 ± 0.1960, p = 0.004) and prolonged eye-closing phase (0.18 ± 0.08 s vs. 0.15 ± 0.07 s, p = 0.032) and opening phase (0.35 ± 0.12 s vs. 0.28 ± 0.14 s, p = 0.015) compared to controls. In addition, Spearman's correlation analysis revealed a negative correlation between incomplete blinks and NIBUT (for first-NIBUT, r = -0.292, p = 0.004; for avg-NIBUT, r = -0.3512, p < 0.001) in children with long-term use of ortho-K.

CONCLUSION: The deep learning system based on U-net and Swim-Transformer achieved optimal performance in the observation of blinking characteristics. Children with long-term use of ortho-K presented an increase in the frequency and rate of incomplete blinks and prolonged eye closing phase and opening phase. The increased frequency of incomplete blinks was associated with decreased tear film stability, indicating the importance of monitoring children's blinking patterns as well as tear film status in clinical follow-up.

PMID:39935789 | PMC:PMC11811098 | DOI:10.3389/fcell.2024.1517240

J R Stat Soc Series B Stat Methodol. 2024 Aug 5;87(1):98-131. doi: 10.1093/jrsssb/qkae082. eCollection 2025 Feb.

ABSTRACT

While fiducial inference was widely considered a big blunder by R.A. Fisher, the goal he initially set-'inferring the uncertainty of model parameters on the basis of observations'-has been continually pursued by many statisticians. To this end, we develop a new statistical inference method called extended Fiducial inference (EFI). The new method achieves the goal of fiducial inference by leveraging advanced statistical computing techniques while remaining scalable for big data. Extended Fiducial inference involves jointly imputing random errors realized in observations using stochastic gradient Markov chain Monte Carlo and estimating the inverse function using a sparse deep neural network (DNN). The consistency of the sparse DNN estimator ensures that the uncertainty embedded in observations is properly propagated to model parameters through the estimated inverse function, thereby validating downstream statistical inference. Compared to frequentist and Bayesian methods, EFI offers significant advantages in parameter estimation and hypothesis testing. Specifically, EFI provides higher fidelity in parameter estimation, especially when outliers are present in the observations; and eliminates the need for theoretical reference distributions in hypothesis testing, thereby automating the statistical inference process. Extended Fiducial inference also provides an innovative framework for semisupervised learning.

PMID:39935678 | PMC:PMC11809222 | DOI:10.1093/jrsssb/qkae082

Front Psychiatry. 2025 Jan 28;16:1508772. doi: 10.3389/fpsyt.2025.1508772. eCollection 2025.

ABSTRACT

INTRODUCTION: Depression is a serious mental health disease. Traditional scale-based depression diagnosis methods often have problems of strong subjectivity and high misdiagnosis rate, so it is particularly important to develop automatic diagnostic tools based on objective indicators.

METHODS: This study proposes a deep learning method that fuses multimodal data to automatically diagnose depression using facial video and audio data. We use spatiotemporal attention module to enhance the extraction of visual features and combine the Graph Convolutional Network (GCN) and the Long and Short Term Memory (LSTM) to analyze the audio features. Through the multi-modal feature fusion, the model can effectively capture different feature patterns related to depression.

RESULTS: We conduct extensive experiments on the publicly available clinical dataset, the Extended Distress Analysis Interview Corpus (E-DAIC). The experimental results show that we achieve robust accuracy on the E-DAIC dataset, with a Mean Absolute Error (MAE) of 3.51 in estimating PHQ-8 scores from recorded interviews.

DISCUSSION: Compared with existing methods, our model shows excellent performance in multi-modal information fusion, which is suitable for early evaluation of depression.

PMID:39935533 | PMC:PMC11811426 | DOI:10.3389/fpsyt.2025.1508772

J Multidiscip Healthc. 2025 Feb 7;18:675-695. doi: 10.2147/JMDH.S493873. eCollection 2025.

ABSTRACT

PURPOSE: Breast cancer is the most common major public health problems of women in the world. Until now, analyzing mammogram images is still the main method used by doctors to diagnose and detect breast cancers. However, this process usually depends on the experience of radiologists and is always very time consuming.

PATIENTS AND METHODS: We propose to introduce deep learning technology into the process for the facilitation of computer-aided diagnosis (CAD), and address the challenges of class imbalance, enhance the detection of small masses and multiple targets, and reduce false positives and negatives in mammogram analysis. Therefore, we adopted and enhanced RetinaNet to detect masses in mammogram images. Specifically, we introduced a novel modification to the network structure, where the feature map M5 is processed by the ReLU function prior to the original convolution kernel. This strategic adjustment was designed to prevent the loss of resolution for small mass features. Additionally, we introduced transfer learning techniques into training process through leveraging pre-trained weights from other RetinaNet applications, and fine-tuned our improved model using the INbreast dataset.

RESULTS: The aforementioned innovations facilitate superior performance of the enhanced RetiaNet model on the public dataset INbreast, as evidenced by a mAP (mean average precision) of 1.0000 and TPR (true positive rate) of 1.00 at 0.00 FPPI (false positive per image) on the INbreast dataset.

CONCLUSION: The experimental results demonstrate that our enhanced RetinaNet model defeats the existing models by having more generalization performance than other published studies, and it can also be applied to other types of patients to assist doctors in making a proper diagnosis.

PMID:39935433 | PMC:PMC11812562 | DOI:10.2147/JMDH.S493873

Sci Rep. 2025 Feb 11;15(1):5093. doi: 10.1038/s41598-025-89096-9.

ABSTRACT

Medical image segmentation plays a crucial role in addressing emerging healthcare challenges. Although several impressive deep learning architectures based on convolutional neural networks (CNNs) and Transformers have recently demonstrated remarkable performance, there is still potential for further performance improvement due to their inherent limitations in capturing feature correlations of input data. To address this issue, this paper proposes a novel encoder-decoder architecture called MedFuseNet that aims to fuse local and global deep feature representations with hybrid attention mechanisms for medical image segmentation. More specifically, the proposed approach contains two branches for feature learning in parallel: one leverages CNNs to learn local correlations of input data, and the other utilizes Swin-Transformer to capture global contextual correlations of input data. For feature fusion and enhancement, the designed hybrid attention mechanisms combine four different attention modules: (1) an atrous spatial pyramid pooling (ASPP) module for the CNN branch, (2) a cross attention module in the encoder for fusing local and global features, (3) an adaptive cross attention (ACA) module in skip connections for further performing fusion, and (4) a squeeze-and-excitation attention (SE-attention) module in the decoder for highlighting informative features. We evaluate our proposed approach on the public ACDC and Synapse datasets, and achieves the average DSC of 89.73% and 78.40%, respectively. Experimental results on these two datasets demonstrate the effectiveness of our proposed approach on medical image segmentation tasks, outperforming other used state-of-the-art approaches.

PMID:39934248 | DOI:10.1038/s41598-025-89096-9

Radiologie (Heidelb). 2025 Feb 11. doi: 10.1007/s00117-025-01422-4. Online ahead of print.

ABSTRACT

BACKGROUND: The rapid development of large language models (LLMs) opens up new possibilities for the automated processing of medical texts. Transforming unstructured radiology reports into structured data is crucial for efficient use in clinical decision support systems, research, and improving patient care.

OBJECTIVES: What are the challenges of transforming natural language radiology reports into structured data using LLMs? Which methods and architectures are promising? How can the quality and reliability of the extracted data be ensured?

MATERIALS AND METHODS: This article examines current research on the application of LLMs in radiological information processing. Various approaches such as rule-based systems, machine learning, and deep learning models, particularly neural network architectures, are analyzed and compared. The focus is on extracting information such as diagnoses, anatomical locations, findings, and measurements.

RESULTS AND CONCLUSION: LLMs show great potential in transforming reports into structured data. In particular, deep learning models trained on large datasets achieve high accuracies. However, challenges remain, such as dealing with ambiguities, abbreviations, and the variability of linguistic expressions. Combining LLMs with domain-specific knowledge, for example, in the form of ontologies, can further improve the performance of the systems. Integrating contextual information and developing robust evaluation metrics are also important research directions.

PMID:39934245 | DOI:10.1007/s00117-025-01422-4

Sci Rep. 2025 Feb 12;15(1):5157. doi: 10.1038/s41598-024-84922-y.

ABSTRACT

Early detection of ocular diseases is vital to preventing severe complications, yet it remains challenging due to the need for skilled specialists, complex imaging processes, and limited resources. Automated solutions are essential to enhance diagnostic precision and support clinical workflows. This study presents a deep learning-based system for automated classification of ocular diseases using the Ocular Disease Intelligent Recognition (ODIR) dataset. The dataset includes 5,000 patient fundus images labeled into eight categories of ocular diseases. Initial experiments utilized transfer learning models such as DenseNet201, EfficientNetB3, and InceptionResNetV2. To optimize computational efficiency, a novel two-level feature selection framework combining Linear Discriminant Analysis (LDA) and advanced neural network classifiers-Deep Neural Networks (DNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BiLSTM)-was introduced. Among the tested approaches, the "Combined Data" strategy utilizing features from all three models achieved the best results, with the BiLSTM classifier attaining 100% accuracy, precision, and recall on the training set, and over 98% performance on the validation set. The LDA-based framework significantly reduced computational complexity while enhancing classification accuracy. The proposed system demonstrates a scalable, efficient solution for ocular disease detection, offering robust support for clinical decision-making. By bridging the gap between clinical demands and technological capabilities, it has the potential to alleviate the workload of ophthalmologists, particularly in resource-constrained settings, and improve patient outcomes globally.

PMID:39934192 | DOI:10.1038/s41598-024-84922-y

Acad Radiol. 2025 Feb 10:S1076-6332(25)00087-X. doi: 10.1016/j.acra.2025.01.032. Online ahead of print.

ABSTRACT

RATIONALE AND OBJECTIVES: Type 2 diabetes is a known risk factor for vascular disease with an impact on the aorta. The aim of this study was to develop a deep learning framework for quantification of aortic phenotypes from magnetic resonance imaging (MRI) and to investigate the association between aortic features and impaired glucose metabolism beyond traditional cardiovascular (CV) risk factors.

MATERIALS AND METHODS: This study used data from the prospective Cooperative Health Research in the Region of Augsburg (KORA) study to develop a deep learning framework for automatic quantification of aortic features (maximum aortic diameter, total volume, length, and width of the aortic arch) derived from MRI. Aortic features were compared between different states of glucose metabolism and tested for associations with impaired glucose metabolism adjusted for traditional CV risk factors (age, sex, height, weight, hypertension, smoking, and lipid panel).

RESULTS: The deep learning framework yielded a high performance for aortic feature quantification with a Dice coefficient of 91.1±0.02. Of 381 participants (58% male, mean age 56 years), 231 (60.6%) had normal blood glucose, 97 (25.5%) had prediabetes, and 53 (13.9%) had diabetes. All aortic features showed a significant increase between different groups of glucose metabolism (p≤0.04). Total aortic length and total aortic volume were associated with impaired glucose metabolism (OR 0.85, 95%CI 0.74-0.96; p=0.01, and OR 0.99, 95%CI 0.98-0.99; p=0.02) independent of CV risk factors.

CONCLUSION: Aortic features showed a glucose level dependent increase from normoglycemic individuals to those with prediabetes and diabetes. Total aortic length and volume were independently and inversely associated with impaired glucose metabolism beyond traditional CV risk factors.

PMID:39934079 | DOI:10.1016/j.acra.2025.01.032

J Prosthet Dent. 2025 Feb 10:S0022-3913(25)00050-2. doi: 10.1016/j.prosdent.2025.01.020. Online ahead of print.

NO ABSTRACT

PMID:39934030 | DOI:10.1016/j.prosdent.2025.01.020

Plant Methods. 2025 Feb 11;21(1):16. doi: 10.1186/s13007-025-01334-3.

ABSTRACT

In the era of machine learning-driven plant imaging, the production of annotated datasets is a very important contribution. In this data paper, a unique annotated dataset of seedling emergence kinetics is proposed. It is composed of almost 70,000 RGB-depth frames and more than 700,000 plant annotations. The dataset is shown valuable for training deep learning models and performing high-throughput phenotyping by imaging. The ability of such models to generalize to several species and outperform the state-of-the-art owing to the delivered dataset is demonstrated. We also discuss how this dataset raises new questions in plant phenotyping.

PMID:39934882 | DOI:10.1186/s13007-025-01334-3

BMC Med Inform Decis Mak. 2025 Feb 11;25(1):73. doi: 10.1186/s12911-025-02895-y.

ABSTRACT

BACKGROUND: Despite recent progress in misinformation detection methods, further investigation is required to develop more robust fact-checking models with particular consideration for the unique challenges of health information sharing. This study aimed to identify the most effective approach for detecting and classifying reliable information versus misinformation health content shared on Twitter/X related to COVID-19.

METHODS: We have used 7 different machine learning/deep learning models. Tweets were collected, processed, labeled, and analyzed using relevant keywords and hashtags, then classified into two distinct datasets: "Trustworthy information" versus "Misinformation", through a labeling process. The cosine similarity metric was employed to address oversampling the minority of the Trustworthy information class, ensuring a more balanced representation of both classes for training and testing purposes. Finally, the performance of the various fact-checking models was analyzed and compared using accuracy, precision, recall, and F1-score ROC curve, and AUC.

RESULTS: For measures of accuracy, precision, F1 score, and recall, the average values of TextConvoNet were found to be 90.28, 90.28, 90.29, and 0.9030, respectively. ROC AUC was 0.901."Trustworthy information" class achieved an accuracy of 85%, precision of 93%, recall of 86%, and F1 score of 89%. These values were higher than other models. Moreover, its performance in the misinformation category was even more impressive, with an accuracy of 94%, precision of 88%, recall of 94%, and F1 score of 91%.

CONCLUSION: This study showed that TextConvoNet was the most effective in detecting and classifying trustworthy information V.S misinformation related to health issues that have been shared on Twitter/X.

PMID:39934858 | DOI:10.1186/s12911-025-02895-y

BMC Med Inform Decis Mak. 2025 Feb 11;25(1):71. doi: 10.1186/s12911-024-02828-1.

ABSTRACT

This study proposes a deep learning-based motion assessment method that integrates the pose estimation algorithm (Keypoint RCNN) with signal processing techniques, demonstrating its reliability and effectiveness.The reliability and validity of this method were also verified.Twenty college students were recruited to pedal a stationary bike. Inertial sensors and a smartphone simultaneously recorded the participants' cycling movement. Keypoint RCNN(KR) algorithm was used to acquire 2D coordinates of the participants' skeletal keypoints from the recorded movement video. Spearman's rank correlation analysis, intraclass correlation coefficient (ICC), error analysis, and t-test were conducted to compare the consistency of data obtained from the two movement capture systems, including the peak frequency of acceleration, transition time point between movement statuses, and the complexity index average (CIA) of the movement status based on multiscale entropy analysis.The KR algorithm showed excellent consistency (ICC1,3=0.988) between the two methods when estimating the peak acceleration frequency. Both peak acceleration frequencies and CIA metrics estimated by the two methods displayed a strong correlation (r > 0.70) and good agreement (ICC2,1>0.750). Additionally, error values were relatively low (MAE = 0.001 and 0.040, MRE = 0.00% and 7.67%). Results of t-tests showed significant differences (p = 0.003 and 0.030) for various acceleration CIAs, indicating our method could distinguish different movement statuses.The KR algorithm also demonstrated excellent intra-session reliability (ICC = 0.988). Acceleration frequency analysis metrics derived from the KR method can accurately identify transitions among movement statuses. Leveraging the KR algorithm and signal processing techniques, the proposed method is designed for individualized motor function evaluation in home or community-based settings.

PMID:39934805 | DOI:10.1186/s12911-024-02828-1

BMC Bioinformatics. 2025 Feb 11;26(1):50. doi: 10.1186/s12859-025-06068-6.

ABSTRACT

BACKGROUND: Piwi-interacting RNAs (piRNAs) are well established for monitoring and protecting the genome from transposons in germline cells. Recently, numerous studies provided evidence that piRNAs also play important roles in regulating mRNA transcript levels. Despite their significant role in regulating cellular RNA levels, the piRNA targeting rules are not well defined, especially in mammals, which poses obstacles to the elucidation of piRNA function.

RESULTS: Given the complexity and current limitation in understanding the mammalian piRNA targeting rules, we designed a deep learning model by selecting appropriate deep learning sub-networks based on the targeting patterns of piRNA inferred from previous experiments. Additionally, to alleviate the problem of insufficient data, a transfer learning approach was employed. Our model achieves a good discriminatory power (Accuracy: 98.5%) in predicting an independent test dataset. Finally, this model was utilized to predict the targets of all mouse and human piRNAs available in the piRNA database.

CONCLUSIONS: In this research, we developed a deep learning framework that significantly advances the prediction of piRNA targets, overcoming the limitations posed by insufficient data and current incomplete targeting rules. The piRNA target prediction network and results can be downloaded from https://github.com/SofiaTianjiaoZhang/piRNATarget .

PMID:39934678 | DOI:10.1186/s12859-025-06068-6

BMC Bioinformatics. 2025 Feb 11;26(1):48. doi: 10.1186/s12859-025-06051-1.

ABSTRACT

BACKGROUND: Alterations of metabolism, including changes in mitochondrial metabolism as well as glutathione (GSH) metabolism are a well appreciated hallmark of many cancers. Mitochondrial GSH (mGSH) transport is a poorly characterized aspect of GSH metabolism, which we investigate in the context of cancer. Existing functional annotation approaches from machine (ML) or deep learning (DL) models based only on protein sequences, were unable to annotate functions in biological contexts.

RESULTS: We develop a flexible ML framework for functional annotation from diverse feature data. This hybrid ML framework leverages cancer cell line multi-omics data and other biological knowledge data as features, to uncover potential genes involved in mGSH metabolism and membrane transport in cancers. This framework achieves strong performance across functional annotation tasks and several cell line and primary tumor cancer samples. For our application, classification models predict the known mGSH transporter SLC25A39 but not SLC25A40 as being highly probably related to mGSH metabolism in cancers. SLC25A10, SLC25A50, and orphan SLC25A24, SLC25A43 are predicted to be associated with mGSH metabolism in multiple biological contexts and structural analysis of these proteins reveal similarities in potential substrate binding regions to the binding residues of SLC25A39.

CONCLUSION: These findings have implications for a better understanding of cancer cell metabolism and novel therapeutic targets with respect to GSH metabolism through potential novel functional annotations of genes. The hybrid ML framework proposed here can be applied to other biological function classifications or multi-omics datasets to generate hypotheses in various biological contexts. Code and a tutorial for generating models and predictions in this framework are available at: https://github.com/lkenn012/mGSH_cancerClassifiers .

PMID:39934670 | DOI:10.1186/s12859-025-06051-1

Sci Rep. 2025 Feb 11;15(1):5080. doi: 10.1038/s41598-025-89752-0.

ABSTRACT

The detection and classification of arrhythmia play a vital role in the diagnosis and management of cardiac disorders. Many deep learning techniques are utilized for arrhythmia classification in current research but only based on ECG data, lacking the mathematical foundations of cardiac electrophysiology. A finite element model (FEM) of the human heart based on the FitzHugh-Nagumo (FHN) model was established for cardiac electrophysiology simulation and the ECG signals were acquired from the FEM results of representative points. Two different kinds of arrhythmia characterized by major anomalies of parameters a and ɛ in the FHN model were simulated, and the synthetic ECG signals were obtained respectively. A multi-objective optimization method based on non-dominated sorting was incorporated into the crayfish optimization algorithm to optimize the key parameters in VMD, then a variational mode decomposition technique for ECG signal processing based on a multi-objective crayfish optimization algorithm (MOCOA-VMD) was proposed, wherein the spectral kurtosis and KL divergence were determined as the indicators for decomposition. The Pareto optimal front was generated by MOCOA and the intrinsic mode functions of VMD with the best combination of K and α were obtained. A deep attention model based on MOCOA-VMD was constructed for ECG signal classification. The ablation study was implemented to verify the effectiveness of the proposed signal decomposition method and deep attention modules. The performance of the model based on MOCOA-VMD achieves the best accuracy of 94.35%, much higher than the model constructed by modules of EEMD, VMD and CNN. Moreover, Bayesian optimization was carried out to fine-tune the hyperparameters batch size, learning rate, epochs, and momentum. After TPE optimization, the deep model's performance achieved a maximum accuracy of 95.91%. The MIT-BIH arrhythmia database was further utilized for model validation, ascertaining its robustness and generalizability. The proposed deep attention modeling and classification strategy can help in arrhythmia signal processing and may offer inspiration for other signal processing fields as well.

PMID:39934416 | DOI:10.1038/s41598-025-89752-0

Sci Rep. 2025 Feb 11;15(1):5117. doi: 10.1038/s41598-025-89768-6.

ABSTRACT

CorneAI, a deep learning model designed for diagnosing cataracts and corneal diseases, was assessed for its impact on ophthalmologists' diagnostic accuracy. In the study, 40 ophthalmologists (20 specialists and 20 residents) classified 100 images, including iPhone 13 Pro photos (50 images) and diffuser slit-lamp photos (50 images), into nine categories (normal condition, infectious keratitis, immunological keratitis, corneal scar, corneal deposit, bullous keratopathy, ocular surface tumor, cataract/intraocular lens opacity, and primary angle-closure glaucoma). The iPhone and slit-lamp images represented the same cases. After initially answering without CorneAI, the same ophthalmologists responded to the same cases with CorneAI 2-4 weeks later. With CorneAI's support, the overall accuracy of ophthalmologists increased significantly from 79.2 to 88.8% (P < 0.001). Specialists' accuracy rose from 82.8 to 90.0%, and residents' from 75.6 to 86.2% (P < 0.001). Smartphone image accuracy improved from 78.7 to 85.5% and slit-lamp image accuracy from 81.2 to 90.6% (both, P < 0.001). In this study, CorneAI's own accuracy was 86%, but its support enhanced ophthalmologists' accuracy beyond the CorneAI's baseline. This study demonstrated that CorneAI, despite being trained on diffuser slit-lamp images, effectively improved diagnostic accuracy, even with smartphone images.

PMID:39934383 | DOI:10.1038/s41598-025-89768-6

Sci Rep. 2025 Feb 11;15(1):5121. doi: 10.1038/s41598-025-88734-6.

ABSTRACT

Stock market stability relies on the shares, investors, and stakeholders' participation and global commodity exchanges. In general, multiple factors influence the stock market stability to ensure profitable returns and commodity transactions. This article presents a contradictory-factor-based stability prediction model using the sigmoid deep learning paradigm. Sigmoid learning identifies the possible stabilizations of different influencing factors toward a profitable stock exchange. In this model, each influencing factor is mapped with the profit outcomes considering the live shares and their exchange value. The stability is predicted using sigmoid and non-sigmoid layers repeatedly until the maximum is reached. This stability is matched with the previous outcomes to predict the consecutive hours of stock market changes. Based on the actual changes and predicted ones, the sigmoid function is altered to accommodate the new range. The non-sigmoid layer remains unchanged in the new changes to improve the prediction precision. Based on the outcomes the deep learning's sigmoid layer is trained to identify abrupt changes in the stock market.

PMID:39934296 | DOI:10.1038/s41598-025-88734-6

Sci Rep. 2025 Feb 11;15(1):5048. doi: 10.1038/s41598-025-89646-1.

ABSTRACT

Deep learning (DL) and explainable artificial intelligence (XAI) have emerged as powerful machine-learning tools to identify complex predictive data patterns in a spatial or temporal domain. Here, we consider the application of DL and XAI to large omic datasets, in order to study biological aging at the molecular level. We develop an advanced multi-view graph-level representation learning (MGRL) framework that integrates prior biological network information, to build molecular aging clocks at cell-type resolution, which we subsequently interpret using XAI. We apply this framework to one of the largest single-cell transcriptomic datasets encompassing over a million immune cells from 981 donors, revealing a ribosomal gene subnetwork, whose expression correlates with age independently of cell-type. Application of the same DL-XAI framework to DNA methylation data of sorted monocytes reveals an epigenetically deregulated inflammatory response pathway whose activity increases with age. We show that the ribosomal module and inflammatory pathways would not have been discovered had we used more standard machine-learning methods. In summary, the computational deep learning framework presented here illustrates how deep learning when combined with explainable AI tools, can reveal novel biological insights into the complex process of aging.

PMID:39934290 | DOI:10.1038/s41598-025-89646-1