Frontiers in Artificial Intelligence最新文献

{"title":"Cryptographic key generation using deep learning with biometric face and finger vein data.","authors":"Tsehayu Gizachew Yirga, Hailu Gizachew Yirga, Eshetie Gizachew Addisu","doi":"10.3389/frai.2025.1545946","DOIUrl":"10.3389/frai.2025.1545946","url":null,"abstract":"<p><p>This research proposes a novel approach to cryptographic key generation using biometric data from face and finger vein modalities enhanced by deep learning techniques. Using pretrained models FaceNet and VGG19 for feature extraction and employing a Siamese Neural Network (SNN), the study demonstrates the integration of multimodal biometrics with fuzzy extractors to create secure and reproducible cryptographic keys. Feature fusion techniques, combined with preprocessing and thresholding, ensure robust feature extraction and conversion to binary formats for key generation. The model demonstrates impressive accuracy with a vector converter, achieving a sigma similarity of 93% and a sigma difference of 64.0%. Evaluation metrics, including False Acceptance Rate (FAR) and False Rejection Rate (FRR), indicate significant improvements, achieving FRR < 3.4% and FAR < 1%, outperforming previous works. Additionally, the adoption of Goppa code-based cryptographic systems ensures post-quantum security. This study not only enhances biometric cryptography's accuracy and resilience but also paves the way for future exploration of quantum-resistant and scalable systems.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1545946"},"PeriodicalIF":3.0,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12069345/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144001154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An efficient method for early Alzheimer's disease detection based on MRI images using deep convolutional neural networks.","authors":"Samia Dardouri","doi":"10.3389/frai.2025.1563016","DOIUrl":"10.3389/frai.2025.1563016","url":null,"abstract":"<p><p>Alzheimer's disease (AD) is a progressive, incurable neurological disorder that leads to a gradual decline in cognitive abilities. Early detection is vital for alleviating symptoms and improving patient quality of life. With a shortage of medical experts, automated diagnostic systems are increasingly crucial in healthcare, reducing the burden on providers and enhancing diagnostic accuracy. AD remains a global health challenge, requiring effective early detection strategies to prevent its progression and facilitate timely intervention. In this study, a deep convolutional neural network (CNN) architecture is proposed for AD classification. The model, consisting of 6,026,324 parameters, uses three distinct convolutional branches with varying lengths and kernel sizes to improve feature extraction. The OASIS dataset used includes 80,000 MRI images sourced from Kaggle, categorized into four classes: non-demented (67,200 images), very mild demented (13,700 images), mild demented (5,200 images), and moderate demented (488 images). To address the dataset imbalance, a data augmentation technique was applied. The proposed model achieved a remarkable 99.68% accuracy in distinguishing between the four stages of Alzheimer's: Non-Dementia, Very Mild Dementia, Mild Dementia, and Moderate Dementia. This high accuracy highlights the model's potential for real-time analysis and early diagnosis of AD, offering a promising tool for healthcare professionals.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1563016"},"PeriodicalIF":3.0,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12069281/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144062460","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Understanding acceptance and resistance toward generative AI technologies: a multi-theoretical framework integrating functional, risk, and sociolegal factors.","authors":"Priyanka Shrivastava","doi":"10.3389/frai.2025.1565927","DOIUrl":"https://doi.org/10.3389/frai.2025.1565927","url":null,"abstract":"<p><p>This study explores the factors influencing college students' acceptance and resistance toward generative AI technologies by integrating three theoretical frameworks: the Technology Acceptance Model (TAM), Protection Motivation Theory (PMT), and Social Exchange Theory (SET). Using data from 407 respondents collected through a structured survey, the study employed Structural Equation Modeling (SEM) to examine how functional factors (perceived usefulness, ease of use, and reliability), risk factors (privacy concerns, data security, and ethical issues), and sociolegal factors (trust in governance and regulatory frameworks) impact user attitudes. Results revealed that functional factors significantly enhanced acceptance while reducing resistance, whereas risk factors amplified resistance and negatively influenced acceptance. Sociolegal factors emerged as critical mediators, mitigating the negative impact of perceived risks and reinforcing the positive effects of functional perceptions. The study responds to prior feedback by offering a more integrated theoretical framework, clearly articulating how TAM, PMT, and SET interact to shape user behavior. It also acknowledges the limitations of using a student sample and discusses the broader applicability of the findings to other demographics, such as professionals and non-academic users. Additionally, the manuscript now highlights demographic diversity, including variations in age, gender, and academic discipline, as relevant to AI adoption patterns. Ethical concerns, including algorithmic bias, data ownership, and the labor market impact of AI, are addressed to offer a more holistic understanding of resistance behavior. Policy implications have been expanded with actionable recommendations such as AI bias mitigation strategies, clearer data ownership protections, and workforce reskilling programs. The study also compares global regulatory frameworks like the GDPR and the U.S. AI Bill of Rights, reinforcing its practical relevance. Furthermore, it emphasizes that user attitudes toward AI are dynamic and likely to evolve, suggesting the need for longitudinal studies to capture behavioral adaptation over time. By bridging theory and practice, this research contributes to the growing discourse on responsible and equitable AI adoption in higher education, offering valuable insights for developers, policymakers, and academic institutions aiming to foster ethical and inclusive technology integration.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1565927"},"PeriodicalIF":3.0,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12066764/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144037760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Layer wise Scaled Gaussian Priors for Markov Chain Monte Carlo Sampled deep Bayesian neural networks.","authors":"Devesh Jawla, John Kelleher","doi":"10.3389/frai.2025.1444891","DOIUrl":"https://doi.org/10.3389/frai.2025.1444891","url":null,"abstract":"<p><p>Previous work has demonstrated that initialization is very important for both fitting a neural network by gradient descent methods, as well as for Variational inference of Bayesian neural networks. In this work we investigate how <i>Layer wise Scaled Gaussian Priors</i> perform with Markov Chain Monte Carlo trained Bayesian neural networks. From our experiments on 8 classifications datasets of various complexity, the results indicate that using <i>Layer wise Scaled Gaussian Priors</i> makes the sampling process more efficient as compared to using an Isotropic Gaussian Prior, an Isotropic Cauchy Prior, or an Isotropic Laplace Prior. We also show that the cold posterior effect does not arise when using a either an Isotropic Gaussian or a layer wise Scaled Prior for small feed forward Bayesian neural networks. Since Bayesian neural networks are becoming popular due to their advantages such as uncertainty estimation, and prevention of over-fitting, this work seeks to provide improvements in the efficiency of Bayesian neural networks learned using Markov Chain Monte Carlo methods.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1444891"},"PeriodicalIF":3.0,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12061901/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144052279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large language models for intelligent RDF knowledge graph construction: results from medical ontology mapping.","authors":"Apostolos Mavridis, Stergios Tegos, Christos Anastasiou, Maria Papoutsoglou, Georgios Meditskos","doi":"10.3389/frai.2025.1546179","DOIUrl":"https://doi.org/10.3389/frai.2025.1546179","url":null,"abstract":"<p><p>The exponential growth of digital data, particularly in specialized domains like healthcare, necessitates advanced knowledge representation and integration techniques. RDF knowledge graphs offer a powerful solution, yet their creation and maintenance, especially for complex medical ontologies like Systematized Nomenclature of Medicine - Clinical Terms (SNOMED CT), remain challenging. Traditional methods often struggle with the scale, heterogeneity, and semantic complexity of medical data. This paper introduces a methodology leveraging the contextual understanding and reasoning capabilities of Large Language Models (LLMs) to automate and enhance medical ontology mapping for Resource Description Framework (RDF) knowledge graph construction. We conduct a comprehensive comparative analysis of six systems-GPT-4o, Claude 3.5 Sonnet v2, Gemini 1.5 Pro, Llama 3.3 70B, DeepSeek R1, and BERTMap-using a novel evaluation framework that combines quantitative metrics (precision, recall, and F1-score) with qualitative assessments of semantic accuracy. Our approach integrates a data preprocessing pipeline with an LLM-powered semantic mapping engine, utilizing BioBERT embeddings and ChromaDB vector database for efficient concept retrieval. Experimental results on a dataset of 108 medical terms demonstrate the superior performance of modern LLMs, particularly GPT-4o, achieving a precision of 93.75% and an F1-score of 96.26%. These findings highlight the potential of LLMs in bridging the gap between structured medical data and semantic knowledge representation, toward more accurate and interoperable medical knowledge graphs.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1546179"},"PeriodicalIF":3.0,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12061982/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143989187","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A novel deep learning technique for multi classify Alzheimer disease: hyperparameter optimization technique.","authors":"A S Elmotelb, Fayroz F Sherif, A S Abohamama, Mahmoud Fakhr, Amr M Abdelatif","doi":"10.3389/frai.2025.1558725","DOIUrl":"https://doi.org/10.3389/frai.2025.1558725","url":null,"abstract":"<p><p>A progressive brain disease that affects memory and cognitive function is Alzheimer's disease (AD). To put therapies in place that potentially slow the progression of AD, early diagnosis and detection are essential. Early detection of these phases enables early activities, which are essential for controlling the disease. To address issues with limited data and computing resources, this work presents a novel deep-learning method based on using a newly proposed hyperparameter optimization method to identify the hyperparameters of ResNet152V2 model for classifying the phases of AD more accurately. The proposed model is compared to state-of-the-art models divided into two categories: transfer learning models and classical models to showcase its effectiveness and efficiency. This comparison is based on four performance metrics: recall, precision, F1 score, and accuracy. According to the experimental results, the proposed method is more efficient and effective in classifying various AD phases.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1558725"},"PeriodicalIF":3.0,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12058654/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144019418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting the Bitcoin's price using AI.","authors":"Gil Cohen, Avishay Aiche","doi":"10.3389/frai.2025.1519805","DOIUrl":"https://doi.org/10.3389/frai.2025.1519805","url":null,"abstract":"<p><p>This study investigates the application of Artificial Intelligence (AI) and Machine Learning (ML) in predicting Bitcoin price movements and developing adaptive investment strategies. An analysis of Bitcoin performance from January 2018 to January 2024 revealed that the AI-driven strategy, leveraging an ensemble of neural networks, achieved a total return of 1640.32%, significantly surpassing the ML-based approach with a return of 304.77% and the traditional B&H strategy at 223.40%. By incorporating predictive analytics and technical indicators, the AI strategy dynamically adjusted its market exposure, enabling it to mitigate losses during downturns and maximize gains during favorable market conditions. These findings underscore the transformative potential of AI in financial markets, particularly in emerging asset classes like cryptocurrencies. Using a broader spectrum of data and employing advanced analytical techniques, AI can provide a more nuanced understanding of market dynamics and investor behavior providing significant implications for portfolio management, risk assessment, and trading system design.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1519805"},"PeriodicalIF":3.0,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12058735/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143986258","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Editorial: The role of conversational AI in higher education.","authors":"Pauldy C J Otermans, Stavros Demetriadis, Deborah Richards","doi":"10.3389/frai.2025.1602037","DOIUrl":"https://doi.org/10.3389/frai.2025.1602037","url":null,"abstract":"","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1602037"},"PeriodicalIF":3.0,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12058729/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144037728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing pre-trained language model by answering natural questions for event extraction.","authors":"Yuxin Zhang, Qing Han","doi":"10.3389/frai.2025.1520290","DOIUrl":"10.3389/frai.2025.1520290","url":null,"abstract":"<p><strong>Introduction: </strong>Event extraction is the task of identifying and extracting structured information about events from unstructured text. However, event extraction remains challenging due to the complexity and diversity of event expressions, as well as the ambiguity and context dependency of language.</p><p><strong>Methods: </strong>In this paper, we propose a new method to improve the precision and recall of event extraction by including topic words related to events and their contexts, directing the model to focus on the relevant information, and filtering the noise.</p><p><strong>Results: </strong>This method was evaluated on the ACE 2005 dataset, achieving an F1-score of 77.27% with significant improvements in both precision and recall.</p><p><strong>Discussion: </strong>Our results show that the use of topic words and question answering techniques can effectively address the challenges faced by event extraction and pave the way for the development of more accurate and robust event extraction systems.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1520290"},"PeriodicalIF":3.0,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12095869/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144128973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep learning for accurate B-line detection and localization in lung ultrasound imaging.","authors":"Nixson Okila, Andrew Katumba, Joyce Nakatumba-Nabende, Cosmas Mwikirize, Sudi Murindanyi, Jonathan Serugunda, Samuel Bugeza, Anthony Oriekot, Juliet Bossa, Eva Nabawanuka","doi":"10.3389/frai.2025.1560523","DOIUrl":"https://doi.org/10.3389/frai.2025.1560523","url":null,"abstract":"<p><strong>Introduction: </strong>Lung ultrasound (LUS) has become an essential imaging modality for assessing various pulmonary conditions, including the presence of B-line artifacts. These artifacts are commonly associated with conditions such as increased extravascular lung water, decompensated heart failure, dialysis-related chronic kidney disease, interstitial lung disease, and COVID-19 pneumonia. Accurate detection of the B-line in LUS images is crucial for effective diagnosis and treatment. However, interpreting LUS is often subject to observer variability, requiring significant expertise and posing challenges in resource-limited settings with few trained professionals.</p><p><strong>Methods: </strong>To address these limitations, deep learning models have been developed for automated B-line detection and localization. This study introduces YOLOv5-PBB and YOLOv8-PBB, two modified models based on YOLOv5 and YOLOv8, respectively, designed for precise and interpretable B-line localization using polygonal bounding boxes (PBBs). YOLOv5-PBB was enhanced by modifying the detection head, loss function, non-maximum suppression, and data loader to enable PBB localization. YOLOv8-PBB was customized to convert segmentation masks into polygonal representations, displaying only boundaries while removing the masks. Additionally, an image preprocessing technique was incorporated into the models to enhance LUS image quality. The models were trained on a diverse dataset from a publicly available repository and Ugandan health facilities.</p><p><strong>Results: </strong>Experimental results showed that YOLOv8-PBB achieved the highest precision (0.947), recall (0.926), and mean average precision (0.957). YOLOv5-PBB, while slightly lower in performance (precision: 0.931, recall: 0.918, mAP: 0.936), had advantages in model size (14 MB vs. 21 MB) and average inference time (33.1 ms vs. 47.7 ms), making it more suitable for real-time applications in low-resource settings.</p><p><strong>Discussion: </strong>The integration of these models into a mobile LUS screening tool provides a promising solution for B-line localization in resource-limited settings, where accessibility to trained professionals may be scarce. The YOLOv5-PBB and YOLOv8-PBB models offer high performance while addressing challenges related to inference speed and model size, making them ideal candidates for mobile deployment in such environments.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"8 ","pages":"1560523"},"PeriodicalIF":3.0,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12053239/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144019156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}