Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548125
Qi Zhuang;Zhengjie Chu;Jun Li
The comments on ice and snow tourism are characterized by high levels of noise, unstructured content, and complex information. However, existing sentiment analysis methods exhibit significant limitations in terms of accuracy and the depth of feature extraction. To address these challenges, this study proposes an intelligent sentiment analysis algorithm based on a multi-model fusion approach: the Improved Dynamic Convolutional and Attention-based Bidirectional Long Short-Term Memory Model (IDCAN-BiLSTM). The aim is to enhance the effectiveness of sentiment analysis for ice and snow tourism reviews. Firstly, the review data is cleaned, denoised, and segmented. High-quality text vector embeddings are then generated using pre-trained Bidirectional Encoder Representations from Transformers (BERT) to capture the deep semantic features of the review text. Subsequently, the IDCAN-BiLSTM model employs a Dynamic Convolutional Neural Network (DCNN) to extract the local features of reviews, thereby increasing sensitivity to specific sentiment words. Following this, a Multi-Head Attention (MHA) mechanism is utilized to focus on key sentiment information within the reviews, effectively addressing the challenges posed by complex and lengthy texts. Finally, the Bidirectional Long Short-Term Memory (BiLSTM) module comprehensively captures the global contextual information in the reviews, improving both the sentiment classification accuracy and the contextual recognition capabilities of the model. Experimental results demonstrate that the IDCAN-BiLSTM model achieves outstanding performance in the sentiment classification of ice and snow tourism reviews, with an accuracy of 92.17% and an F1 score of 0.93. These results significantly surpass those of traditional sentiment analysis methods. In particular, the model shows superior performance in the sentiment classification of long review texts, effectively enhancing the accuracy and granularity of sentiment recognition through dynamic convolution and the self-attention mechanism. Moreover, the model distinguishes sentiment tendencies across different user groups regarding their experiences in ice and snow tourism. This capability provides valuable data support for optimizing services and enabling precision marketing strategies in the ice and snow tourism sector.
{"title":"Text Analysis of Digital Commentary on Ice and Snow Tourism Based on Artificial Intelligence and Long Short-Term Memory Neural Network","authors":"Qi Zhuang;Zhengjie Chu;Jun Li","doi":"10.1109/ACCESS.2025.3548125","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548125","url":null,"abstract":"The comments on ice and snow tourism are characterized by high levels of noise, unstructured content, and complex information. However, existing sentiment analysis methods exhibit significant limitations in terms of accuracy and the depth of feature extraction. To address these challenges, this study proposes an intelligent sentiment analysis algorithm based on a multi-model fusion approach: the Improved Dynamic Convolutional and Attention-based Bidirectional Long Short-Term Memory Model (IDCAN-BiLSTM). The aim is to enhance the effectiveness of sentiment analysis for ice and snow tourism reviews. Firstly, the review data is cleaned, denoised, and segmented. High-quality text vector embeddings are then generated using pre-trained Bidirectional Encoder Representations from Transformers (BERT) to capture the deep semantic features of the review text. Subsequently, the IDCAN-BiLSTM model employs a Dynamic Convolutional Neural Network (DCNN) to extract the local features of reviews, thereby increasing sensitivity to specific sentiment words. Following this, a Multi-Head Attention (MHA) mechanism is utilized to focus on key sentiment information within the reviews, effectively addressing the challenges posed by complex and lengthy texts. Finally, the Bidirectional Long Short-Term Memory (BiLSTM) module comprehensively captures the global contextual information in the reviews, improving both the sentiment classification accuracy and the contextual recognition capabilities of the model. Experimental results demonstrate that the IDCAN-BiLSTM model achieves outstanding performance in the sentiment classification of ice and snow tourism reviews, with an accuracy of 92.17% and an F1 score of 0.93. These results significantly surpass those of traditional sentiment analysis methods. In particular, the model shows superior performance in the sentiment classification of long review texts, effectively enhancing the accuracy and granularity of sentiment recognition through dynamic convolution and the self-attention mechanism. Moreover, the model distinguishes sentiment tendencies across different user groups regarding their experiences in ice and snow tourism. This capability provides valuable data support for optimizing services and enabling precision marketing strategies in the ice and snow tourism sector.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"41259-41269"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910133","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548031
Eyob Solomon Getachew;Beakal Gizachew Assefa
The Collatz conjecture, which posits that any positive integer will eventually reach 1 through a specific iterative process, is a classic unsolved problem in mathematics. This research focuses on designing an efficient algorithm to compute the stopping time of numbers in the Collatz sequence, achieving significant computational improvements. By leveraging structural patterns in the Collatz tree, the proposed algorithm minimizes redundant operations and optimizes computational steps. Unlike prior methods, it efficiently handles extremely large numbers without requiring advanced techniques such as memoization or parallelization. Experimental evaluations confirm computational efficiency improvements of approximately 28% over state-of-the-art methods. These findings underscore the algorithm’s scalability and robustness, providing a foundation for future large-scale verification of the conjecture and potential applications in computational mathematics.
{"title":"Efficient Computation of Collatz Sequence Stopping Times: A Novel Algorithmic Approach","authors":"Eyob Solomon Getachew;Beakal Gizachew Assefa","doi":"10.1109/ACCESS.2025.3548031","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548031","url":null,"abstract":"The Collatz conjecture, which posits that any positive integer will eventually reach 1 through a specific iterative process, is a classic unsolved problem in mathematics. This research focuses on designing an efficient algorithm to compute the stopping time of numbers in the Collatz sequence, achieving significant computational improvements. By leveraging structural patterns in the Collatz tree, the proposed algorithm minimizes redundant operations and optimizes computational steps. Unlike prior methods, it efficiently handles extremely large numbers without requiring advanced techniques such as memoization or parallelization. Experimental evaluations confirm computational efficiency improvements of approximately 28% over state-of-the-art methods. These findings underscore the algorithm’s scalability and robustness, providing a foundation for future large-scale verification of the conjecture and potential applications in computational mathematics.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"41210-41220"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910185","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548318
Yeonwoo Kim;Bosung Jeon;Jonghyuk Park;Woo Young Choi
Hardware spiking neural networks (SNNs) are gaining increasing attention as promising next-generation computing approaches owing to their parallel computing and energy-efficient nature. Recent SNNs have achieved excellent performance by introducing artificial neural networks (ANN) to the SNN conversion training method. However, SNNs face significant challenges when implemented on hardware platforms. These challenges include achieving low energy consumption, difficulties in minimizing conversion error, and representing ANN functions other than the rectified linear unit activation function. This paper proposes mixed-signal complementary metal-oxide-semiconductor time-frame integrate-and-fire (TIF) neuron circuits and corresponding SNNs to achieve high accuracy with low energy consumption. TIF neurons integrate synaptic inputs during a time-frame and generate output spikes under synchronization signals based on the integrated input. When building SNNs with TIF neurons, bias and max-pooling can be effectively implemented, which is difficult to achieve using conventional SNNs with integrate-and-fire (IF) neuron circuits. In addition, TIF neurons support pipelining operations, further enhancing the inference throughput of the system. Simulation results with the CIFAR-10 dataset show that SNNs with TIF neurons achieve 0.48 %p higher classification accuracy while consuming 92 % lower neuron energy than conventional IF-based SNNs.
{"title":"Time-Frame Integrate-and-Fire Neuron Circuit for Low Energy Inference Hardware Spiking Neural Networks","authors":"Yeonwoo Kim;Bosung Jeon;Jonghyuk Park;Woo Young Choi","doi":"10.1109/ACCESS.2025.3548318","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548318","url":null,"abstract":"Hardware spiking neural networks (SNNs) are gaining increasing attention as promising next-generation computing approaches owing to their parallel computing and energy-efficient nature. Recent SNNs have achieved excellent performance by introducing artificial neural networks (ANN) to the SNN conversion training method. However, SNNs face significant challenges when implemented on hardware platforms. These challenges include achieving low energy consumption, difficulties in minimizing conversion error, and representing ANN functions other than the rectified linear unit activation function. This paper proposes mixed-signal complementary metal-oxide-semiconductor time-frame integrate-and-fire (TIF) neuron circuits and corresponding SNNs to achieve high accuracy with low energy consumption. TIF neurons integrate synaptic inputs during a time-frame and generate output spikes under synchronization signals based on the integrated input. When building SNNs with TIF neurons, bias and max-pooling can be effectively implemented, which is difficult to achieve using conventional SNNs with integrate-and-fire (IF) neuron circuits. In addition, TIF neurons support pipelining operations, further enhancing the inference throughput of the system. Simulation results with the CIFAR-10 dataset show that SNNs with TIF neurons achieve 0.48 %p higher classification accuracy while consuming 92 % lower neuron energy than conventional IF-based SNNs.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"43728-43740"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910109","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alzheimer’s disease (AD) presents a significant global health challenge, necessitating accurate and early prediction methods for effective intervention and treatment planning. In this work, a novel approach to meta-learning for the prediction of AD is proposed, which leverages the combined power of neural processes (NPs) and transformer architectures. We introduce a framework that integrates NPs with a transformer encoder to model the complex temporal dependencies inherent in longitudinal health data, where our model learns to capture subtle patterns and variations indicative of disease progression. The novelty of our approach lies in the fusion of NPs, renowned for their ability to model stochastic processes, with transformer architectures, known for their ability to capture long-range dependencies. This combination enables our model to effectively adapt to individual patient trajectories and generalize across diverse populations, enhancing its predictive performance and robustness. We trained our proposed model with the Alzheimer’s Disease Prediction Of Longitudinal Evolution dataset (TADPOLE), which contains three classes: Cognitively Normal (CN), Mild Cognitive Impairment (MCI), and AD. The experimental results demonstrate that the proposed model enhances the prediction of these models in terms of mAUC, Recall, and Precision by $0.937pm 0.014$ , $0.920pm 0.010$ , and $0.923pm 0.009$ , respectively. These findings prove the efficacy of the proposed framework in accurately predicting the progression of AD, highlighting its potential for early detection and personalized treatment strategies.
{"title":"A Novel Approach to the Prediction of Alzheimer’s Disease Progression by Leveraging Neural Processes and a Transformer Encoder Model","authors":"Emad Al-Anbari;Hossein Karshenas;Bijan Shoushtarian","doi":"10.1109/ACCESS.2025.3548173","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548173","url":null,"abstract":"Alzheimer’s disease (AD) presents a significant global health challenge, necessitating accurate and early prediction methods for effective intervention and treatment planning. In this work, a novel approach to meta-learning for the prediction of AD is proposed, which leverages the combined power of neural processes (NPs) and transformer architectures. We introduce a framework that integrates NPs with a transformer encoder to model the complex temporal dependencies inherent in longitudinal health data, where our model learns to capture subtle patterns and variations indicative of disease progression. The novelty of our approach lies in the fusion of NPs, renowned for their ability to model stochastic processes, with transformer architectures, known for their ability to capture long-range dependencies. This combination enables our model to effectively adapt to individual patient trajectories and generalize across diverse populations, enhancing its predictive performance and robustness. We trained our proposed model with the Alzheimer’s Disease Prediction Of Longitudinal Evolution dataset (TADPOLE), which contains three classes: Cognitively Normal (CN), Mild Cognitive Impairment (MCI), and AD. The experimental results demonstrate that the proposed model enhances the prediction of these models in terms of mAUC, Recall, and Precision by <inline-formula> <tex-math>$0.937pm 0.014$ </tex-math></inline-formula>, <inline-formula> <tex-math>$0.920pm 0.010$ </tex-math></inline-formula>, and <inline-formula> <tex-math>$0.923pm 0.009$ </tex-math></inline-formula>, respectively. These findings prove the efficacy of the proposed framework in accurately predicting the progression of AD, highlighting its potential for early detection and personalized treatment strategies.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"44607-44619"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910173","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548128
Abdalrahman Alblwi;Saleh Makkawy;Kenneth E. Barner
The Denoising Diffusion Probabilistic Model (DDPM) has gained significant attention for its powerful image generation and segmentation capabilities, particularly in biomedical applications where accuracy is critical. In breast cancer detection, ultrasound imaging is widely used due to its safety, affordability, and non-ionizing nature. However, the inherent challenges of ultrasound data, such as noise and artifacts, make accurate tumor segmentation difficult, often leading to misdiagnosis. We propose a novel Deep Denoising Probabilistic Diffusion Model (D-DDPM) designed to enhance tumor segmentation in breast ultrasound images to address these limitations. Our model incorporates a nested U-Net architecture with Residual U-blocks (RSU), significantly improving feature learning and segmentation precision. In addition to performing segmentation, D-DDPM generates synthetic data, augmenting existing real datasets to improve data size with a diverse range of high-quality samples. We validated D-DDPM on several breast ultrasound datasets, comparing its performance to state-of-the-art methods. The proposed D-DDPM achieves a Dice score improvement of 2.26%, 4.24%, and 5% over the runner-up model, demonstrating superior performance on all BUS datasets. Both qualitative and quantitative results demonstrate the ability of D-DDPM to deliver more accurate and reliable segmentation results, offering promising potential to improve clinical decision-making in cancer diagnosis.
{"title":"D-DDPM: Deep Denoising Diffusion Probabilistic Models for Lesion Segmentation and Data Generation in Ultrasound Imaging","authors":"Abdalrahman Alblwi;Saleh Makkawy;Kenneth E. Barner","doi":"10.1109/ACCESS.2025.3548128","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548128","url":null,"abstract":"The Denoising Diffusion Probabilistic Model (DDPM) has gained significant attention for its powerful image generation and segmentation capabilities, particularly in biomedical applications where accuracy is critical. In breast cancer detection, ultrasound imaging is widely used due to its safety, affordability, and non-ionizing nature. However, the inherent challenges of ultrasound data, such as noise and artifacts, make accurate tumor segmentation difficult, often leading to misdiagnosis. We propose a novel Deep Denoising Probabilistic Diffusion Model (D-DDPM) designed to enhance tumor segmentation in breast ultrasound images to address these limitations. Our model incorporates a nested U-Net architecture with Residual U-blocks (RSU), significantly improving feature learning and segmentation precision. In addition to performing segmentation, D-DDPM generates synthetic data, augmenting existing real datasets to improve data size with a diverse range of high-quality samples. We validated D-DDPM on several breast ultrasound datasets, comparing its performance to state-of-the-art methods. The proposed D-DDPM achieves a Dice score improvement of 2.26%, 4.24%, and 5% over the runner-up model, demonstrating superior performance on all BUS datasets. Both qualitative and quantitative results demonstrate the ability of D-DDPM to deliver more accurate and reliable segmentation results, offering promising potential to improve clinical decision-making in cancer diagnosis.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"41194-41209"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910103","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621585","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548424
Praveen B. Choppala;Ramoni Adeogun
The auxiliary particle filter, which is the popular extension of the standard bootstrap particle filter, is known to assist in drawing particles from regions of high probability mass of the posterior density by leveraging the incoming measurement information in the sampling process. The filter accomplishes this by looking ahead in time to determine those particles that become important when propagated forward, retract, and then propagate those particles forward in time. The key problem with this approach is that a particle determined to be important may not fall in regions of importance when actually propagated forward, either because of a large diffusion of the state transition kernel and/or a highly informative measurement, thus defeating the entire purpose of the filter. This problem leads to degeneracy. This paper proposes a method of sampling a multitude of particles for each particle to make such a decision. The key idea here is to use multiple disturbances, instead of one as does the auxiliary particle filter, as lookahead means to guide particles to regions of high probability in the posterior probability density. Through evaluation, we show that the proposed idea overcomes the said problem and exhibits less degeneracy and high tracking accuracy.
{"title":"Auxiliary Particle Filtering With Multitudinous Lookahead Sampling for Accurate Target Tracking","authors":"Praveen B. Choppala;Ramoni Adeogun","doi":"10.1109/ACCESS.2025.3548424","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548424","url":null,"abstract":"The auxiliary particle filter, which is the popular extension of the standard bootstrap particle filter, is known to assist in drawing particles from regions of high probability mass of the posterior density by leveraging the incoming measurement information in the sampling process. The filter accomplishes this by looking ahead in time to determine those particles that become important when propagated forward, retract, and then propagate those particles forward in time. The key problem with this approach is that a particle determined to be important may not fall in regions of importance when actually propagated forward, either because of a large diffusion of the state transition kernel and/or a highly informative measurement, thus defeating the entire purpose of the filter. This problem leads to degeneracy. This paper proposes a method of sampling a multitude of particles for each particle to make such a decision. The key idea here is to use multiple disturbances, instead of one as does the auxiliary particle filter, as lookahead means to guide particles to regions of high probability in the posterior probability density. Through evaluation, we show that the proposed idea overcomes the said problem and exhibits less degeneracy and high tracking accuracy.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"42874-42886"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910112","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143594327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Prediction of myocardial infarction (MI) is crucial for early intervention and treatment. Machine learning has increasingly been applied in the realm of disease prediction. This study explores the feasibility of utilizing easily obtainable heart rate (HR) and respiratory rate (RR) data collected during nocturnal sleep, in conjunction with clinical characteristics and medical domain knowledge, to predict MI. Data for this investigation were sourced from the Sleep Heart Health Study (SHHS) program in the United States, which was categorized into MI and non-MI groups based on the occurrence or absence of MI during follow-up, involving a total of 488 participants. Multiple features related to HR and RR were extracted and integrated with clinical features; four algorithms—MLP, SVM, XGBoost, and CNN—were employed for model construction. The findings indicated that the MLP model exhibited superior performance, achieving an accuracy rate 71.1%. Furthermore, three medical rules age, HR, and RR were incorporated into the MLP model to mitigate the limitations of small sample sizes. The experiments demonstrate that the model’s accuracy reaches its optimal level by combining the age rule, improving to 73.1%. The findings indicate that leveraging non-cardiac electrophysiological data obtained during sleep alongside medical domain knowledge can significantly enhance the accuracy of early predictions regarding cardiac MI while offering novel insights for its prevention and diagnosis.
{"title":"Prediction of Myocardial Infarction Based on Non-ECG Sleep Data Combined With Domain Knowledge","authors":"Changyun Li;Yonghan Zhao;Qihui Mo;Zhibing Wang;Xi Xu","doi":"10.1109/ACCESS.2025.3548118","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548118","url":null,"abstract":"Prediction of myocardial infarction (MI) is crucial for early intervention and treatment. Machine learning has increasingly been applied in the realm of disease prediction. This study explores the feasibility of utilizing easily obtainable heart rate (HR) and respiratory rate (RR) data collected during nocturnal sleep, in conjunction with clinical characteristics and medical domain knowledge, to predict MI. Data for this investigation were sourced from the Sleep Heart Health Study (SHHS) program in the United States, which was categorized into MI and non-MI groups based on the occurrence or absence of MI during follow-up, involving a total of 488 participants. Multiple features related to HR and RR were extracted and integrated with clinical features; four algorithms—MLP, SVM, XGBoost, and CNN—were employed for model construction. The findings indicated that the MLP model exhibited superior performance, achieving an accuracy rate 71.1%. Furthermore, three medical rules age, HR, and RR were incorporated into the MLP model to mitigate the limitations of small sample sizes. The experiments demonstrate that the model’s accuracy reaches its optimal level by combining the age rule, improving to 73.1%. The findings indicate that leveraging non-cardiac electrophysiological data obtained during sleep alongside medical domain knowledge can significantly enhance the accuracy of early predictions regarding cardiac MI while offering novel insights for its prevention and diagnosis.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"42262-42271"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910191","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143621603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Stunting, a critical health issue affecting child growth and development, is prevalent in developing countries and is characterized by significantly reduced height for age. Traditional height measurement methods often require physical contact, which can lead to measurement inaccuracies and potential discomfort for infants. This study introduces a contactless method for measuring infant height using advanced computer vision techniques and the MediaPipe Pose library. By detecting key body points and applying Euclidean distance calculations, the proposed approach offers precise height estimation. Validation uses baby dolls (38 cm and 49 cm) and real infants (n =12) under varying body postures and lighting conditions. A fixed-size green mat (100 cm) was used as a reference for converting pixel distances into actual measurements. The method achieved an average accuracy of 99.76% for the 38 cm doll and 99.67% for the 49 cm doll. For real infants, the system demonstrated an average accuracy of 98.48%. This confirms that the system performs effectively in measuring infant height, even under conditions of non-ideal body posture. Furthermore, these results suggest that the proposed system is an effective and practical alternative for infant height measurements, supporting the early detection of stunting in diverse real-world settings.
{"title":"Contactless Infant Height Measurement for Enhanced Early Detection of Stunting Using Computer Vision Techniques","authors":"Risfendra;Aripriharta;Suherman;Gheri Febri Ananda;Dwi Sudarno Putra;Fahmi","doi":"10.1109/ACCESS.2025.3548159","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548159","url":null,"abstract":"Stunting, a critical health issue affecting child growth and development, is prevalent in developing countries and is characterized by significantly reduced height for age. Traditional height measurement methods often require physical contact, which can lead to measurement inaccuracies and potential discomfort for infants. This study introduces a contactless method for measuring infant height using advanced computer vision techniques and the MediaPipe Pose library. By detecting key body points and applying Euclidean distance calculations, the proposed approach offers precise height estimation. Validation uses baby dolls (38 cm and 49 cm) and real infants (n =12) under varying body postures and lighting conditions. A fixed-size green mat (100 cm) was used as a reference for converting pixel distances into actual measurements. The method achieved an average accuracy of 99.76% for the 38 cm doll and 99.67% for the 49 cm doll. For real infants, the system demonstrated an average accuracy of 98.48%. This confirms that the system performs effectively in measuring infant height, even under conditions of non-ideal body posture. Furthermore, these results suggest that the proposed system is an effective and practical alternative for infant height measurements, supporting the early detection of stunting in diverse real-world settings.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"42364-42376"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10910102","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143594278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-05DOI: 10.1109/ACCESS.2025.3548591
Ramteja Sajja;Vinay Pursnani;Yusuf Sermet;Ibrahim Demir
Floodplain management is critical for mitigating flood risks and safeguarding communities. The FloodPlain Manager (FPM) certification is essential for professionals in this field, but current preparation methods often fall short in providing comprehensive, accessible, and engaging study resources. This research introduces a novel AI-assisted educational tool designed specifically for FPM certification preparation and training process. Leveraging advanced natural language processing and machine learning techniques, this tool offers personalized learning experiences, interactive question-and-answer sessions, and real-time feedback to aspiring floodplain managers. The system architecture integrates certification-specific content through a sophisticated document parsing process, ensuring relevance and accuracy. Evaluation of the tool, conducted through text similarity analysis, demonstrates its effectiveness in preparing candidates for the FPM certification exam. With 91.7% accuracy for open-ended questions and 95.12% for multiple-choice questions, the tool offers a personalized learning experience through dynamic flashcards and adaptive quizzes, highlighting its potential to enhance vocational training and exam readiness. This study underscores the transformative role of AI in professional education and suggests future directions for expanding the tool’s capabilities and application to other certifications.
{"title":"AI-Assisted Educational Framework for Floodplain Manager Certification: Enhancing Vocational Education and Training Through Personalized Learning","authors":"Ramteja Sajja;Vinay Pursnani;Yusuf Sermet;Ibrahim Demir","doi":"10.1109/ACCESS.2025.3548591","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548591","url":null,"abstract":"Floodplain management is critical for mitigating flood risks and safeguarding communities. The FloodPlain Manager (FPM) certification is essential for professionals in this field, but current preparation methods often fall short in providing comprehensive, accessible, and engaging study resources. This research introduces a novel AI-assisted educational tool designed specifically for FPM certification preparation and training process. Leveraging advanced natural language processing and machine learning techniques, this tool offers personalized learning experiences, interactive question-and-answer sessions, and real-time feedback to aspiring floodplain managers. The system architecture integrates certification-specific content through a sophisticated document parsing process, ensuring relevance and accuracy. Evaluation of the tool, conducted through text similarity analysis, demonstrates its effectiveness in preparing candidates for the FPM certification exam. With 91.7% accuracy for open-ended questions and 95.12% for multiple-choice questions, the tool offers a personalized learning experience through dynamic flashcards and adaptive quizzes, highlighting its potential to enhance vocational training and exam readiness. This study underscores the transformative role of AI in professional education and suggests future directions for expanding the tool’s capabilities and application to other certifications.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"42401-42413"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10912484","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143594411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This note considers the active secure event-triggered control(ETC) problem of hybrid power system under DoS attack. A combination of load frequency control (LFC) and virtual inertia control (VIC) is adopted to deal with the influence of uncertainty and lower inertia induced by renewable energy like wind and solar power. To active defend DoS attack interrupting communication of measurement and control, an interval secure event-triggered mechanism (ISETM) is proposed under software defined network (SDN). Both a triggering transmission and a secure triggering interval are generated where the triggering packet is transmitted over SDN data plane and the secure triggering interval is sent to SDN control plane regulating SDN cybersecurity mechanism. Under ISETM, multi-area hybrid power system is modeled by a delay system with two triggering conditions. Furthermore, interval secure event-triggered LFC-VIC of hybrid power system is formulated by a $H_{infty } $ control problem. A sufficient criterion of hybrid power system with the prescribed $H_{infty } $ performance level is derived by using Lyapunov-Krasovskii functional method. A co-designed method of ISETM and LFC-VIC gains is given by linear matrix inequalities (LMIs). Finally, a two-area hybrid power system is simulated to verify the validness of the proposed interval secure event-triggered control (ISETC) method.
{"title":"Interval Secure Event-Triggered Control of Hybrid Power System Under DoS Attack","authors":"Dashuang Chong;Tongshu Si;Zihao Cheng;Feng Yang;Jigang Liu;Zongwang Lv","doi":"10.1109/ACCESS.2025.3548455","DOIUrl":"https://doi.org/10.1109/ACCESS.2025.3548455","url":null,"abstract":"This note considers the active secure event-triggered control(ETC) problem of hybrid power system under DoS attack. A combination of load frequency control (LFC) and virtual inertia control (VIC) is adopted to deal with the influence of uncertainty and lower inertia induced by renewable energy like wind and solar power. To active defend DoS attack interrupting communication of measurement and control, an interval secure event-triggered mechanism (ISETM) is proposed under software defined network (SDN). Both a triggering transmission and a secure triggering interval are generated where the triggering packet is transmitted over SDN data plane and the secure triggering interval is sent to SDN control plane regulating SDN cybersecurity mechanism. Under ISETM, multi-area hybrid power system is modeled by a delay system with two triggering conditions. Furthermore, interval secure event-triggered LFC-VIC of hybrid power system is formulated by a <inline-formula> <tex-math>$H_{infty } $ </tex-math></inline-formula> control problem. A sufficient criterion of hybrid power system with the prescribed <inline-formula> <tex-math>$H_{infty } $ </tex-math></inline-formula> performance level is derived by using Lyapunov-Krasovskii functional method. A co-designed method of ISETM and LFC-VIC gains is given by linear matrix inequalities (LMIs). Finally, a two-area hybrid power system is simulated to verify the validness of the proposed interval secure event-triggered control (ISETC) method.","PeriodicalId":13079,"journal":{"name":"IEEE Access","volume":"13 ","pages":"42574-42586"},"PeriodicalIF":3.4,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10912443","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143601966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}