Sci Rep. 2025 Jan 25;15(1):3204. doi: 10.1038/s41598-025-87450-5.

ABSTRACT

In order to solve the limitations of flipped classroom in personalized teaching and interactive effect improvement, this paper designs a new model of flipped classroom in colleges and universities based on Virtual Reality (VR) by combining the algorithm of Contrastive Language-Image Pre-Training (CLIP). Through cross-modal data fusion, the model deeply combines students' operation behavior with teaching content, and improves teaching effect through intelligent feedback mechanism. The test data shows that the similarity between video and image modes reaches 0.89, which indicates that different modal information can be effectively integrated to ensure the semantic consistency and intuitive understanding of teaching content. The minimum Kullback-Leibler (KL) divergence is 0.12, which ensures the stability of data distribution and avoids information loss. The accuracy of automatically generating feedback reaches 93.72%, which significantly improves the efficiency of personalized learning guidance. In the adaptability test of virtual scene, the frequency of scene adjustment is 2.5 times/minute, and the consistency score is stable above 8.6, ensuring the consistency of teaching goals under complex interaction. This paper aims to enhance personalized learning experience, improve teaching efficiency and autonomous learning effect through VR technology and intelligent feedback, and promote the innovation of interactive teaching mode.

PMID:39863690 | DOI:10.1038/s41598-025-87450-5

NPJ Precis Oncol. 2025 Jan 26;9(1):27. doi: 10.1038/s41698-025-00808-w.

ABSTRACT

Patients with High-Grade Serous Ovarian Cancer (HGSOC) exhibit varied responses to treatment, with 20-30% showing de novo resistance to platinum-based chemotherapy. While hematoxylin-eosin (H&E)-stained pathological slides are used for routine diagnosis of cancer type, they may also contain diagnostically useful information about treatment response. Our study demonstrates that combining H&E-stained whole slide images (WSIs) with proteomic signatures using a multimodal deep learning framework significantly improves the prediction of platinum response in both discovery and validation cohorts. This method outperforms the Homologous Recombination Deficiency (HRD) score in predicting platinum response and overall patient survival. Our study suggests that histology and proteomics contain complementary information about biological processes determining response to first line platinum treatment in HGSOC. This integrative approach has the potential to improve personalized treatment and provide insights into the therapeutic vulnerabilities of HGSOC.

PMID:39863682 | DOI:10.1038/s41698-025-00808-w

Sci Data. 2025 Jan 25;12(1):149. doi: 10.1038/s41597-025-04447-5.

ABSTRACT

With climate extremes' rising frequency and intensity, robust analytical tools are crucial to predict their impacts on terrestrial ecosystems. Machine learning techniques show promise but require well-structured, high-quality, and curated analysis-ready datasets. Earth observation datasets comprehensively monitor ecosystem dynamics and responses to climatic extremes, yet the data complexity can challenge the effectiveness of machine learning models. Despite recent progress in deep learning to ecosystem monitoring, there is a need for datasets specifically designed to analyse compound heatwave and drought extreme impact. Here, we introduce the DeepExtremeCubes database, tailored to map around these extremes, focusing on persistent natural vegetation. It comprises over 40,000 globally sampled small data cubes (i.e. minicubes), with a spatial coverage of 2.5 by 2.5 km. Each minicube includes (i) Sentinel-2 L2A images, (ii) ERA5-Land variables and generated extreme event cube covering 2016 to 2022, and (iii) ancillary land cover and topography maps. The paper aims to (1) streamline data accessibility, structuring, pre-processing, and enhance scientific reproducibility, and (2) facilitate biosphere dynamics forecasting in response to compound extremes.

PMID:39863624 | DOI:10.1038/s41597-025-04447-5

Nat Commun. 2025 Jan 25;16(1):1031. doi: 10.1038/s41467-025-56412-w.

ABSTRACT

Compute-in-memory based on resistive random-access memory has emerged as a promising technology for accelerating neural networks on edge devices. It can reduce frequent data transfers and improve energy efficiency. However, the nonvolatile nature of resistive memory raises concerns that stored weights can be easily extracted during computation. To address this challenge, we propose RePACK, a threefold data protection scheme that safeguards neural network input, weight, and structural information. It utilizes a bipartite-sort coding scheme to store data with a fully on-chip physical unclonable function. Experimental results demonstrate the effectiveness of increasing enumeration complexity to 5.77 × 1075 for a 128-column compute-in-memory core. We further implement and evaluate a RePACK computing system on a 40 nm resistive memory compute-in-memory chip. This work represents a step towards developing safe, robust, and efficient edge neural network accelerators. It potentially serves as the hardware infrastructure for edge devices in federated learning or other systems.

PMID:39863590 | DOI:10.1038/s41467-025-56412-w

Am J Orthod Dentofacial Orthop. 2025 Feb;167(2):242-249. doi: 10.1016/j.ajodo.2024.09.006.

ABSTRACT

INTRODUCTION: This study aimed to assess the precision of an open-source, clinician-trained, and user-friendly convolutional neural network-based model for automatically segmenting the mandible.

METHODS: A total of 55 cone-beam computed tomography scans that met the inclusion criteria were collected and divided into test and training groups. The MONAI (Medical Open Network for Artificial Intelligence) Label active learning tool extension was used to train the automatic model. To assess the model's performance, 15 cone-beam computed tomography scans from the test group were inputted into the model. The ground truth was obtained from manual segmentation data. Metrics including the Dice similarity coefficient, Hausdorff 95%, precision, recall, and segmentation times were calculated. In addition, surface deviations and volumetric differences between the automated and manual segmentation results were analyzed.

RESULTS: The automated model showed a high level of similarity to the manual segmentation results, with a mean Dice similarity coefficient of 0.926 ± 0.014. The Hausdorff distance was 1.358 ± 0.466 mm, whereas the mean recall and precision values were 0.941 ± 0.028 and 0.941 ± 0.022, respectively. There were no statistically significant differences in the arithmetic mean of the surface deviation for the entire mandible and 11 different anatomic regions. In terms of volumetric comparisons, the difference between the 2 groups was 1.62 mm³, which was not statistically significant.

CONCLUSIONS: The automated model was found to be suitable for clinical use, demonstrating a high degree of agreement with the reference manual method. Clinicians can use open-source software to develop custom automated segmentation models tailored to their specific needs.

PMID:39863342 | DOI:10.1016/j.ajodo.2024.09.006

Health Informatics J. 2025 Jan-Mar;31(1):14604582251315594. doi: 10.1177/14604582251315594.

ABSTRACT

Purpose: In the context of Chinese clinical texts, this paper aims to propose a deep learning algorithm based on Bidirectional Encoder Representation from Transformers (BERT) to identify privacy information and to verify the feasibility of our method for privacy protection in the Chinese clinical context. Methods: We collected and double-annotated 33,017 discharge summaries from 151 medical institutions on a municipal regional health information platform, developed a BERT-based Bidirectional Long Short-Term Memory Model (BiLSTM) and Conditional Random Field (CRF) model, and tested the performance of privacy identification on the dataset. To explore the performance of different substructures of the neural network, we created five additional baseline models and evaluated the impact of different models on performance. Results: Based on the annotated data, the BERT model pre-trained with the medical corpus showed a significant performance improvement to the BiLSTM-CRF model with a micro-recall of 0.979 and an F1 value of 0.976, which indicates that the model has promising performance in identifying private information in Chinese clinical texts. Conclusions: The BERT-based BiLSTM-CRF model excels in identifying privacy information in Chinese clinical texts, and the application of this model is very effective in protecting patient privacy and facilitating data sharing.

PMID:39862116 | DOI:10.1177/14604582251315594

Viruses. 2024 Dec 25;17(1):14. doi: 10.3390/v17010014.

ABSTRACT

Human respiratory syncytial virus (RSV) remains a significant global health threat, particularly for vulnerable populations. Despite extensive research, effective antiviral therapies are still limited. To address this urgent need, we present AVP-GPT2, a deep-learning model that significantly outperforms its predecessor, AVP-GPT, in designing and screening antiviral peptides. Trained on a significantly expanded dataset, AVP-GPT2 employs a transformer-based architecture to generate diverse peptide sequences. A multi-modal screening approach, incorporating Star-Transformer and Vision Transformer, enables accurate prediction of antiviral activity and toxicity, leading to the identification of potent and safe candidates. SHAP analysis further enhances interpretability by explaining the underlying mechanisms of peptide activity. Our in vitro experiments confirmed the antiviral efficacy of peptides generated by AVP-GPT2, with some exhibiting EC50 values as low as 0.01 μM and CC50 values > 30 μM. This represents a substantial improvement over AVP-GPT and traditional methods. AVP-GPT2 has the potential to significantly impact antiviral drug discovery by accelerating the identification of novel therapeutic agents. Future research will explore its application to other viral targets and its integration into existing drug development pipelines.

PMID:39861804 | DOI:10.3390/v17010014

Sensors (Basel). 2025 Jan 20;25(2):589. doi: 10.3390/s25020589.

ABSTRACT

With the rapid development of AI algorithms and computational power, object recognition based on deep learning frameworks has become a major research direction in computer vision. UAVs equipped with object detection systems are increasingly used in fields like smart transportation, disaster warning, and emergency rescue. However, due to factors such as the environment, lighting, altitude, and angle, UAV images face challenges like small object sizes, high object density, and significant background interference, making object detection tasks difficult. To address these issues, we use YOLOv8s as the basic framework and introduce a multi-level feature fusion algorithm. Additionally, we design an attention mechanism that links distant pixels to improve small object feature extraction. To address missed detections and inaccurate localization, we replace the detection head with a dynamic head, allowing the model to route objects to the appropriate head for final output. We also introduce Slideloss to improve the model's learning of difficult samples and ShapeIoU to better account for the shape and scale of bounding boxes. Experiments on datasets like VisDrone2019 show that our method improves accuracy by nearly 10% and recall by about 11% compared to the baseline. Additionally, on the AI-TODv1.5 dataset, our method improves the mAP50 from 38.8 to 45.2.

PMID:39860960 | DOI:10.3390/s25020589

Sensors (Basel). 2025 Jan 20;25(2):580. doi: 10.3390/s25020580.

ABSTRACT

The Internet of Things (IoT) has emerged as a crucial element in everyday life. The IoT environment is currently facing significant security concerns due to the numerous problems related to its architecture and supporting technology. In order to guarantee the complete security of the IoT, it is important to deal with these challenges. This study centers on employing deep learning methodologies to detect attacks. In general, this research aims to improve the performance of existing deep learning models. To mitigate data imbalances and enhance learning outcomes, the synthetic minority over-sampling technique (SMOTE) is employed. Our approach contributes to a multistage feature extraction process where autoencoders (AEs) are used initially to extract robust features from unstructured data on the model architecture's left side. Following this, long short-term memory (LSTM) networks on the right analyze these features to recognize temporal patterns indicative of abnormal behavior. The extracted and temporally refined features are inputted into convolutional neural networks (CNNs) for final classification. This structured arrangement harnesses the distinct capabilities of each model to process and classify IoT security data effectively. Our framework is specifically designed to address various attacks, including denial of service (DoS) and Mirai attacks, which are particularly harmful to IoT systems. Unlike conventional intrusion detection systems (IDSs) that may employ a singular model or simple feature extraction methods, our multistage approach provides more comprehensive analysis and utilization of data, enhancing detection capabilities and accuracy in identifying complex cyber threats in IoT environments. This research highlights the potential benefits that can be gained by applying deep learning methods to improve the effectiveness of IDSs in IoT security. The results obtained indicate a potential improvement for enhancing security measures and mitigating emerging threats.

PMID:39860948 | DOI:10.3390/s25020580

Sensors (Basel). 2025 Jan 19;25(2):561. doi: 10.3390/s25020561.

ABSTRACT

Real-time and accurate traffic forecasting aids in traffic planning and design and helps to alleviate congestion. Addressing the negative impacts of partial data loss in traffic forecasting, and the challenge of simultaneously capturing short-term fluctuations and long-term trends, this paper presents a traffic forecasting model, D-MGDCN-CLSTM, based on Multi-Graph Gated Dilated Convolution and Conv-LSTM. The model uses the DTWN algorithm to fill in missing data. To better capture the dual characteristics of short-term fluctuations and long-term trends in traffic, the model employs the DWT for multi-scale decomposition to obtain approximation and detail coefficients. The Conv-LSTM processes the approximation coefficients to capture the long-term characteristics of the time series, while the multiple layers of the MGDCN process the detail coefficients to capture short-term fluctuations. The outputs of the two branches are then merged to produce the forecast results. The model is tested against 10 algorithms using the PeMSD7(M) and PeMSD7(L) datasets, improving MAE, RMSE, and ACC by an average of 1.38% and 13.89%, 1% and 1.24%, and 5.92% and 1%, respectively. Ablation experiments, parameter impact analysis, and visual analysis all demonstrate the superiority of our decompositional approach in handling the dual characteristics of traffic data.

PMID:39860932 | DOI:10.3390/s25020561

Sensors (Basel). 2025 Jan 19;25(2):555. doi: 10.3390/s25020555.

ABSTRACT

This study investigates mental fatigue in sports activities by leveraging deep learning techniques, deviating from the conventional use of heart rate variability (HRV) feature analysis found in previous research. The study utilizes a hybrid deep neural network model, which integrates Residual Networks (ResNet) and Bidirectional Long Short-Term Memory (Bi-LSTM) for feature extraction, and a transformer for feature fusion. The model achieves an impressive accuracy of 95.29% in identifying fatigue from original ECG data, 2D spectral characteristics and physiological information of subjects. In comparison to traditional methods, such as Support Vector Machines (SVMs) and Random Forests (RFs), as well as other deep learning methods, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM), the proposed approach demonstrates significantly improved experimental outcomes. Overall, this study offers a promising solution for accurately recognizing fatigue through the analysis of physiological signals, with potential applications in sports and physical fitness training contexts.

PMID:39860925 | DOI:10.3390/s25020555

Sensors (Basel). 2025 Jan 19;25(2):554. doi: 10.3390/s25020554.

ABSTRACT

Predicting the Remaining Useful Life (RUL) is vital for ensuring the reliability and safety of equipment and components. This study introduces a novel method for predicting RUL that utilizes the Convolutional Block Attention Module (CBAM) to address the problem that Convolutional Neural Networks (CNNs) do not effectively leverage data channel features and spatial features in residual life prediction. Firstly, Fast Fourier Transform (FFT) is applied to convert the data into the frequency domain. The resulting frequency domain data is then used as input to the convolutional neural network for feature extraction; Then, the weights of channel features and spatial features are assigned to the extracted features by CBAM, and the weighted features are then input into the Long Short-Term Memory (LSTM) network to learn temporal features. Finally, the effectiveness of the proposed model is verified using the PHM2012 bearing dataset. Compared to several existing RUL prediction methods, the mean squared error, mean absolute error, and root mean squared error of the proposed method in this paper are reduced by 53%, 16.87%, and 31.68%, respectively, which verifies the superiority of the method. Meanwhile, the experimental results demonstrate that the proposed method achieves good RUL prediction accuracy across various failure modes.

PMID:39860924 | DOI:10.3390/s25020554

Sensors (Basel). 2025 Jan 18;25(2):547. doi: 10.3390/s25020547.

ABSTRACT

Tea bud localization detection not only ensures tea quality, improves picking efficiency, and advances intelligent harvesting, but also fosters tea industry upgrades and enhances economic benefits. To solve the problem of the high computational complexity of deep learning detection models, we developed the Tea Bud DSCF-YOLOv8n (TBF-YOLOv8n)lightweight detection model. Improvement of the Cross Stage Partial Bottleneck Module with Two Convolutions(C2f) module via efficient Distributed Shift Convolution (DSConv) yields the C2f module with DSConv(DSCf)module, which reduces the model's size. Additionally, the coordinate attention (CA) mechanism is incorporated to mitigate interference from irrelevant factors, thereby improving mean accuracy. Furthermore, the SIOU_Loss (SCYLLA-IOU_Loss) function and the Dynamic Sample(DySample)up-sampling operator are implemented to accelerate convergence and enhance both average precision and detection accuracy. The experimental results show that compared to the YOLOv8n model, the TBF-YOLOv8n model has a 3.7% increase in accuracy, a 1.1% increase in average accuracy, a 44.4% reduction in gigabit floating point operations (GFLOPs), and a 13.4% reduction in the total number of parameters included in the model. In comparison experiments with a variety of lightweight detection models, the TBF-YOLOv8n still performs well in terms of detection accuracy while remaining more lightweight. In conclusion, the TBF-YOLOv8n model achieves a commendable balance between efficiency and precision, offering valuable insights for advancing intelligent tea bud harvesting technologies.

PMID:39860916 | DOI:10.3390/s25020547

Sensors (Basel). 2025 Jan 18;25(2):545. doi: 10.3390/s25020545.

ABSTRACT

With the proliferation of mobile terminals and the rapid growth of network applications, fine-grained traffic identification has become increasingly challenging. Methods based on machine learning and deep learning have achieved remarkable results, but they heavily rely on the distribution of training data, which makes them ineffective in handling unseen samples. In this paper, we propose AG-ZSL, a zero-shot learning framework based on traffic behavior and attribute representations for general encrypted traffic classification. AG-ZSL primarily learns two mapping functions: one that captures traffic behavior embeddings from burst-based traffic interaction graphs, and the other that learns attribute embeddings from traffic attribute descriptions. Then, the framework minimizes the distance between these embeddings within the shared feature space. The gradient rejection algorithm and K-Nearest Neighbors are introduced to implement a two-stage method for general traffic classification. Experimental results on IoT datasets demonstrate that AG-ZSL achieves exceptional performance in classifying both known and unknown traffic, highlighting its potential for enhancing secure and efficient traffic management at the network edge.

PMID:39860913 | DOI:10.3390/s25020545

Sensors (Basel). 2025 Jan 17;25(2):534. doi: 10.3390/s25020534.

ABSTRACT

The issue of obstacle avoidance and safety for visually impaired individuals has been a major topic of research. However, complex street environments still pose significant challenges for blind obstacle detection systems. Existing solutions often fail to provide real-time, accurate obstacle avoidance decisions. In this study, we propose a blind obstacle detection system based on the PC-CS-YOLO model. The system improves the backbone network by adopting the partial convolutional feed-forward network (PCFN) to reduce computational redundancy. Additionally, to enhance the network's robustness in multi-scale feature fusion, we introduce the Cross-Scale Attention Fusion (CSAF) mechanism, which integrates features from different sensory domains to achieve superior performance. Compared to state-of-the-art networks, our system shows improvements of 2.0%, 3.9%, and 1.5% in precision, recall, and mAP50, respectively. When evaluated on a GPU, the inference speed is 20.6 ms, which is 15.3 ms faster than YOLO11, meeting the real-time requirements for blind obstacle avoidance systems.

PMID:39860905 | DOI:10.3390/s25020534

Biomed Phys Eng Express. 2025 Jan 24. doi: 10.1088/2057-1976/adae14. Online ahead of print.

ABSTRACT

Traumatic injury remains a leading cause of death worldwide, with traumatic bleeding being one of its most critical and fatal consequences. The use of whole-body computed tomography (WBCT) in trauma management has rapidly expanded. However, interpreting WBCT images within the limited time available before treatment is particularly challenging for acute care physicians. Our group has previously developed an automated bleeding detection method in WBCT images. However, further reduction of false positives (FPs) is necessary for clinical application. To address this issue, we propose a novel automated detection for traumatic bleeding in CT images using deep learning and multi-organ segmentation; Methods: The proposed method integrates a three-dimensional U-Net# model for bleeding detection with an FP reduction approach based on multi-organ segmentation. The multi-organ segmentation method targets the bone, kidney, and vascular regions, where FPs are primarily found during the bleeding detection process. We evaluated the proposed method using a dataset of delayed-phase contrast-enhanced trauma CT images collected from four institutions; Results: Our method detected 70.0% of bleedings with 76.2 FPs/case. The processing time for our method was 6.3 ± 1.4 min. Compared with our previous ap-proach, the proposed method significantly reduced the number of FPs while maintaining detection sensitivity.

PMID:39854772 | DOI:10.1088/2057-1976/adae14

PLoS One. 2025 Jan 24;20(1):e0312177. doi: 10.1371/journal.pone.0312177. eCollection 2025.

ABSTRACT

In human activity-recognition scenarios, including head and entire body pose and orientations, recognizing the pose and direction of a pedestrian is considered a complex problem. A person may be traveling in one sideway while focusing his attention on another side. It is occasionally desirable to analyze such orientation estimates using computer-vision tools for automated analysis of pedestrian behavior and intention. This article uses a deep-learning method to demonstrate the pedestrian full-body pose estimation approach. A deep-learning-based pre-trained supervised model multi-branched deep learning pose net (MBDLP-Net) is proposed for estimation and classification. For full-body pose and orientation estimation, three independent datasets, an extensive dataset for body orientation (BDBO), PKU-Reid, and TUD Multiview Pedestrians, are used. Independently, the proposed technique is trained on dataset CIFAR-100 with 100 classes. The proposed approach is meticulously tested using publicly accessible BDBO, PKU-Reid, and TUD datasets. The results show that the mean accuracy for full-body pose estimation with BDBO and PKU-Reid is 0.95%, and with TUD multiview pedestrians is 0.97%. The performance results show that the proposed technique efficiently distinguishes full-body poses and orientations in various configurations. The efficacy of the provided approach is compared with existing pretrained, robust, and state-of-the-art methodologies, providing a comprehensive understanding of its advantages.

PMID:39854382 | DOI:10.1371/journal.pone.0312177

PLoS One. 2025 Jan 24;20(1):e0312363. doi: 10.1371/journal.pone.0312363. eCollection 2025.

ABSTRACT

The traditional method of corn quality detection relies heavily on the subjective judgment of inspectors and suffers from a high error rate. To address these issues, this study employs the Swin Transformer as an enhanced base model, integrating machine vision and deep learning techniques for corn quality assessment. Initially, images of high-quality, moldy, and broken corn were collected. After preprocessing, a total of 20,152 valid images were obtained for the experimental samples. The network then extracts both shallow and deep features from these maize images, which are subsequently fused. Concurrently, the extracted features undergo further processing through a specially designed convolutional block. The fused features, combined with those processed by the convolutional module, are fed into an attention layer. This attention layer assigns weights to the features, facilitating accurate final classification. Experimental results demonstrate that the MC-Swin Transformer model proposed in this paper significantly outperforms traditional convolutional neural network models in key metrics such as accuracy, precision, recall, and F1 score, achieving a recognition accuracy rate of 99.89%. Thus, the network effectively and efficiently classifies different corn qualities. This study not only offers a novel perspective and technical approach to corn quality detection but also holds significant implications for the advancement of smart agriculture.

PMID:39854315 | DOI:10.1371/journal.pone.0312363

Bioinformatics. 2025 Jan 24:btaf035. doi: 10.1093/bioinformatics/btaf035. Online ahead of print.

ABSTRACT

MOTIVATION: In silico functional annotation of proteins is crucial to narrowing the sequencing-accelerated gap in our understanding of protein activities. Numerous function annotation methods exist, and their ranks have been growing, particularly so with the recent deep learning-based developments. However, it is unclear if these tools are truly predictive. As we are not aware of any methods that can identify new terms in functional ontologies, we ask if they can, at least, identify molecular functions of proteins that are non-homologous to or far-removed from known protein families.

RESULTS: Here, we explore the potential and limitations of the existing methods in predicting molecular functions of thousands of such proteins. Lacking the "ground truth" functional annotations, we transformed the assessment of function prediction into evaluation of functional similarity of protein pairs that likely share function but are unlike any of the currently functionally annotated sequences. Notably, our approach transcends the limitations of functional annotation vocabularies, providing a means to assess different-ontology annotation methods. We find that most existing methods are limited to identifying functional similarity of homologous sequences and fail to predict function of proteins lacking reference. Curiously, despite their seemingly unlimited by-homology scope, deep learning methods also have trouble capturing the functional signal encoded in protein sequence. We believe that our work will inspire the development of a new generation of methods that push boundaries and promote exploration and discovery in the molecular function domain.

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PMID:39854283 | DOI:10.1093/bioinformatics/btaf035

Transl Vis Sci Technol. 2025 Jan 2;14(1):27. doi: 10.1167/tvst.14.1.27.

ABSTRACT

PURPOSE: The integration of artificial intelligence (AI), particularly deep learning (DL), with optical coherence tomography (OCT) offers significant opportunities in the diagnosis and management of glaucoma. This article explores the application of various DL models in enhancing OCT capabilities and addresses the challenges associated with their clinical implementation.

METHODS: A review of articles utilizing DL models was conducted, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), autoencoders, and large language models (LLMs). Key developments and practical applications of these models in OCT image analysis were emphasized, particularly in the context of enhancing image quality, glaucoma diagnosis, and monitoring progression.

RESULTS: CNNs excel in segmenting retinal layers and detecting glaucomatous damage, whereas RNNs are effective in analyzing sequential OCT scans for disease progression. GANs enhance image quality and data augmentation, and autoencoders facilitate advanced feature extraction. LLMs show promise in integrating textual and visual data for comprehensive diagnostic assessments. Despite these advancements, challenges such as data availability, variability, potential biases, and the need for extensive validation persist.

CONCLUSIONS: DL models are reshaping glaucoma management by enhancing OCT's diagnostic capabilities. However, the successful translation into clinical practice requires addressing major challenges related to data variability, biases, fairness, and model validation to ensure accurate and reliable patient care.

TRANSLATIONAL RELEVANCE: This review bridges the gap between basic research and clinical care by demonstrating how AI, particularly DL models, can markedly enhance OCT's clinical utility in diagnosis, monitoring, and prediction, moving toward more individualized, personalized, and precise treatment strategies.

PMID:39854198 | DOI:10.1167/tvst.14.1.27