Article
Neurosciences
Yuwen Ruan, Xiang Chen, Xu Zhang, Xun Chen
Summary: This study explores the feasibility of using principal component analysis (PCA) algorithm to separate gesture pattern-related signals from noise and proposes a PCA-based PPG signal processing scheme to improve gesture recognition accuracy. Experimental results show that PCA decomposition effectively separates relevant signals from noise, and the proposed scheme is particularly effective for finger-related gestures.
FRONTIERS IN NEUROSCIENCE
(2022)
Article
Agriculture, Multidisciplinary
Zhen Zhang, Jun Zhou, Zhenghong Yan, Kai Wang, Jiamin Mao, Zizhen Jiang
Summary: Hardness recognition of fruits and vegetables based on tactile array information was proposed in this study, with PCA-KNN and PCA-SVM algorithms achieving accuracy rates of 90.03% and 94.27% respectively. This method allowed the robot to stably grasp products without damage, as verified by an online grabbing recognition experiment with a 90% accuracy rate.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Computer Science, Information Systems
Alex X. Wang, Stefanka S. Chukova, Binh P. Nguyen
Summary: k-nearest neighbors (k-NN) is a well-known classification algorithm that is widely used in different domains. We proposed the Centroid Displacement-based k-NN algorithm to address the issue of class determination in standard k-NN algorithm. Our experimental results demonstrate that our algorithm is able to enhance the classification performance of the standard k-NN algorithm and its variants and also improve the computational efficiency.
INFORMATION SCIENCES
(2023)
Article
Chemistry, Analytical
Azrina Abd Aziz, Lila Iznita Izhar, Vijanth Sagayan Asirvadam, Tong Boon Tang, Azimah Ajam, Zaid Omar, Sobri Muda
Summary: This study focuses on the challenging task of identifying collateral vessels from CBCT images of acute stroke patients, proposing a technique to objectively distinguish collateral vessels from non-collateral vessels using various feature extraction methods and classifiers. The results demonstrate that the classifiers achieve promising classification accuracy above 90% and are able to effectively detect collateral and non-collateral vessels from the images.
Article
Chemistry, Analytical
Jee S. Ra, Tianning Li, Yan Li
Summary: This research aims to develop an efficient and accurate algorithm for predicting epileptic seizures by optimizing channel selection based on feature extraction and classification techniques. Results show a significant improvement in prediction accuracy, sensitivity, and specificity using selected channels compared to testing with all channels. Tailored channel selection proves to be a robust way to optimize seizure prediction.
Article
Computer Science, Artificial Intelligence
Arushi Agarwal, Purushottam Sharma, Mohammed Alshehri, Ahmed A. Mohamed, Osama Alfarraj
Summary: This study improves the accuracy and reduces processing time of IDS algorithms by comparing different classification machine learning algorithms to find the best-suited algorithm to learn suspicious network activity patterns. Data gathered from feature set comparison is used as input to train the system for future intrusion behavior prediction and analysis.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Artificial Intelligence
Anita Maria da Rocha Fernandes, Mateus Junior Cassaniga, Bianka Tallita Passos, Eros Comunello, Stefano Frizzo Stefenon, Valderi Reis Quietinho Leithardt
Summary: The paper proposes the use of machine learning techniques to detect cracks and potholes in images of paved roads, and through experiments comparing and classifying images, it proves that this technique is feasible for locating faults in highways.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2023)
Article
Medicine, General & Internal
Mohamed Sraitih, Younes Jabrane, Amir Hajjam El Hassani
Summary: This study investigates an automatic ECG myocardial infarction detection system and proposes a new approach to evaluate its performance in classifying myocardial infarction under different noise types. The results show that the machine learning models used perform well in classifying myocardial infarction in the presence of different types of noise. These models can be used as detection tools for myocardial infarction in challenging environments.
JOURNAL OF CLINICAL MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Konstantinos Kalpakis
Summary: Classification and regression algorithms based on k-nearest neighbors (kNN) are popular in machine learning due to their performance, flexibility, interpretability, and computational efficiency. However, existing kNN algorithms do not consider the weights associated with training samples, which limits their performance and flexibility. We propose a new weighted kNN algorithm that incorporates weights for training samples and demonstrates its effectiveness in mitigating covariate shift and improving performance compared to benchmark approaches.
Article
Materials Science, Multidisciplinary
Ying Chen, Haochen Peng, Yuzhu Liu
Summary: Identification based on LIBS and machine learning is important for reducing the risk of using low-quality or inappropriate charcoal. This study used different types of charcoal samples for detection and achieved classification accuracy of 96.0% and 97.3% using optimized methods. The results indicate that LIBS combined with machine learning provides a new and effective method for charcoal detection and classification.
JOURNAL OF LASER APPLICATIONS
(2021)
Article
Computer Science, Theory & Methods
Xiaojia Song, Tao Xie, Stephen Fischer
Summary: In this paper, two kNN kernels are implemented on FPGA using high-level synthesis (HLS) and two data access reduction methods - low-precision data representation (LPDR) and principal component analysis based filtering (PCAF). The experimental results show that these two kernels greatly improve the performance by reducing external memory-accesses, compared to other implementations.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2022)
Article
Agronomy
Priya Brata Bhoi, Veeresh S. Wali, Deepak Kumar Swain, Kalpana Sharma, Akash Kumar Bhoi, Manlio Bacco, Paolo Barsocchi
Summary: This research highlights the variation in technical efficiency of paddy cultivation across different states in India, with different inputs affecting the yield. Utilizing machine learning algorithms, the study found that the random forest model achieved the highest mean accuracy of 0.80 for predicting farmers' efficiency class. The study demonstrates a technique for classifying and predicting a farmer's efficiency group based on input parameters for each state.
Article
Computer Science, Information Systems
Shanshan Liu, Pedro Reviriego, Jose Alberto Hernandez, Fabrizio Lombardi
Summary: This paper explores how to provide protection and error tolerance for classifiers by exploiting the algorithmic properties, applied to the k Nearest Neighbors classifier, and proposes a time-based modular redundancy scheme to reduce the number of re-computations needed effectively.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING
(2021)
Article
Neurosciences
Francesco Ferracuti, Sabrina Iarlori, Zahra Mansour, Andrea Monteriu, Camillo Porcaro
Summary: The development of brain-computer interface (BCI) technology allows for controlling external devices using brain signals, improving the quality of life for people with motor disabilities. This study implemented a data-driven automatic channels selection method to improve the classification accuracy of brain signals.
Article
Materials Science, Multidisciplinary
Mark Kirk, Yoshinori Hashimoto, Akiyoshi Nomoto
Summary: Considerable effort has been made by researchers to produce codified embrittlement trend curves for nuclear reactor pressure vessels. This paper explores the insights made possible by a more localized approach to data fitting using nearest-neighbor techniques. The objective is to evaluate the accuracy and precision of nearest neighbor predictions of transition temperature shift, using a comprehensive international surveillance database.
JOURNAL OF NUCLEAR MATERIALS
(2022)
Article
Computer Science, Information Systems
Allam Jaya Prakash, Kiran Kumar Patro, Saunak Samantray, Pawel Plawiak, Mohamed Hammad
Summary: An electrocardiogram (ECG) is a unique representation of a person's identity, and its rhythm and shape are completely different from person to person, making it difficult to clone or tamper with. This paper proposes a beat-based template matching deep learning technique to address the challenges in traditional techniques, such as noise, feature extraction, and system performance.
Article
Engineering, Biomedical
Asmaa Maher, Saeed Mian Qaisar, N. Salankar, Feng Jiang, Ryszard Tadeusiewicz, Pawel Plawiak, Ahmed A. Abd El-Latif, Mohamed Hammad
Summary: The Brain-computer interface (BCI) is used to enhance human capabilities. Researchers have started focusing on the Electroencephalogram (EEG) and Functional Near-Infrared Spectroscopy (fNIRS) based hybrid-BCI (hBCI) due to the development of artificial intelligence (AI) algorithms. A novel EEG-fNIRS based hBCI system is devised using non-linear features mining and ensemble learning (EL) approach, achieving high accuracy, F1-score, and sensitivity.
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING
(2023)
Article
Geochemistry & Geophysics
P. Homola, V. Marchenko, A. Napolitano, R. Damian, R. Guzik, D. Alvarez-Castillo, S. Stuglik, O. Ruimi, O. Skorenok, J. Zamora-Saa, J. M. Vaquero, T. Wibig, M. Knap, K. Dziadkowiec, M. Karpiel, O. Sushchov, J. W. Mietelski, K. Gorzkiewicz, N. Zabari, K. Almeida Cheminant, B. Idzkowski, T. Bulik, G. Bhatta, N. Budnev, R. Kaminski, M. V. Medvedev, K. Kozak, O. Bar, L. Bibrzycki, M. Bielewicz, M. Frontczak, P. Kovacs, B. Lozowski, J. Miszczyk, M. Niedzwiecki, L. del Peral, M. Piekarczyk, M. D. Rodriguez Frias, K. Rzecki, K. Smelcerz, T. Sosnicki, J. Stasielak, A. A. Tursunov
Summary: For the first time, we have discovered a correlation between the average variation in secondary cosmic ray detection rates and global seismic activity, with a time lag of approximately two weeks. The significance of this effect varies in a periodic manner resembling the undecennial solar cycle, with a phase shift of around three years, exceeding 6 sigma at local maxima. The precursor characteristics of these correlations suggest the possibility of developing an early warning system against earthquakes.
JOURNAL OF ATMOSPHERIC AND SOLAR-TERRESTRIAL PHYSICS
(2023)
Article
Chemistry, Analytical
Shimaa Saber, Souham Meshoul, Khalid Amin, Pawel Plawiak, Mohamed Hammad
Summary: This paper presents a novel approach for person re-identification by introducing a multi-part feature network that combines the position attention module (PAM) and the efficient channel attention (ECA). The proposed method outperforms existing state-of-the-art methods on publicly available person re-ID datasets. The results demonstrate the effectiveness and potential of the suggested method in computer vision applications.
Article
Chemistry, Multidisciplinary
Mateusz Baran, Zbislaw Tabor, Krzysztof Rzecki, Przemyslaw Ziaja, Tomasz Szumlak, Kamila Kalecinska, Jakub Michczynski, Bartlomiej Rachwal, Michael P. R. Waligorski, David Sarrut
Summary: The accurate calculation of dose distribution in external photon beam therapy is crucial for successful oncology treatment using a medical linear accelerator. Monte Carlo simulation is currently the most accurate method for this purpose, but it is computationally extensive. In this study, we propose a method of generating phase spaces in medical linear accelerators using artificial intelligence models. Our results show that the differences between dose distributions obtained using generative-adversarial-network-based and Monte-Carlo-based phase spaces are very close, indicating the potential of this method for faster and accurate dose delivery calculation.
APPLIED SCIENCES-BASEL
(2023)
Review
Oncology
Navneet Melarkode, Kathiravan Srinivasan, Saeed Mian Qaisar, Pawel Plawiak
Summary: This research aims to explore the use of deep learning and machine learning techniques for diagnosing skin cancer. It discusses the challenges and future directions in this field, as well as compares widely used datasets and review papers on skin cancer diagnosis using AI. The authors of this study aim to establish a benchmark for further research in this field and address the limitations and benefits of historical approaches.
Article
Mathematics
Sergey Bushelenkov, Alexander Paramonov, Ammar Muthanna, Ahmed A. Abd El-Latif, Andrey Koucheryavy, Osama Alfarraj, Pawel Plawiak, Abdelhamied A. Ateya
Summary: This article introduces a new network model for IoT based on a multi-story building structure. The model adopts a regular, cubic lattice-like structure for placing network nodes, leading to an equation for the signal-to-noise ratio (SNR). The study also establishes the correlation between traffic density, network density, and SNR. Furthermore, the article explores the potential of percolation theory in characterizing network functionality. The findings offer a novel approach to network design and planning, enabling the selection of a network topology that meets criteria and requirements while ensuring connectivity and enhancing efficiency. The developed analytical apparatus provides valuable insights into the properties of the network and its applicability to specific conditions.
Article
Chemistry, Analytical
Asif Rahim, Yanru Zhong, Tariq Ahmad, Sadique Ahmad, Pawel Plawiak, Mohamed Hammad
Summary: This study explores the application of deep learning models for anomaly detection and face recognition in IoT devices within the context of smart homes. The LR-HGBC-CNN model consistently outperformed the others in both anomaly detection and face recognition tasks, achieving high accuracy, precision, recall, F1 score, and AUC-ROC values. The study suggests that deep learning approaches have the potential to enhance security and privacy in smart homes, but further research is needed to evaluate their generalizability and address deployment challenges.
Article
Chemistry, Analytical
Erana Veerappa Dinesh Subramaniam, Kathiravan Srinivasan, Saeed Mian Qaisar, Pawel Plawiak
Summary: The emergence of the Internet of Medical Things (IoMT) has enabled remote patient diagnosis and treatment using mobile-device-collected data. However, concerns about patient privacy arise when using traditional AI systems in this context. To address this issue, we propose a privacy-enhanced approach for illness diagnosis within the IoMT framework, improving IoT network performance and ensuring data confidentiality using various techniques. Our approach shows substantial enhancements in network performance metrics compared with previous works, emphasizing the effectiveness of our method in enhancing IoT network interoperability and protection, and improving patient care and diagnostic capabilities.
Article
Chemistry, Analytical
Achraf Daoui, Mohamed Yamni, Torki Altameem, Musheer Ahmad, Mohamed Hammad, Pawel Plawiak, Ryszard Tadeusiewicz, Ahmed A. Abd El-Latif
Summary: This paper introduces a scheme called AuCFSR to ensure the authenticity of color face images and recover tampered areas. The scheme uses a two-dimensional hyperchaotic system to embed authentication and recovery data, producing high-security and high-quality output images. Experimental results demonstrate that AuCFSR outperforms other schemes in terms of tamper detection accuracy, security level, and visual quality.
Review
Chemistry, Analytical
Manar Osama, Abdelhamied A. Ateya, Mohammed S. Sayed, Mohamed Hammad, Pawel Plawiak, Ahmed A. Abd El-Latif, Rania A. Elsayed
Summary: Healthcare 4.0 is a recent e-health paradigm associated with Industry 4.0, providing precision medicine based on patient's characteristics and enabling telemedicine. It is driven by technologies like 5G, wearable devices, AI, edge computing, and IoT. It is important for modern societies, especially during pandemics.
Article
Chemistry, Analytical
Mohamed Yamni, Achraf Daoui, Pawel Plawiak, Haokun Mao, Osama Alfarraj, Ahmed A. Abd El-Latif
Summary: In this study, an integer and reversible version of the Krawtchouk transform (IRKT) is proposed to avoid rounding errors in floating-point processing. Based on the IRKT, a reliable 3D reversible data hiding (RDH) algorithm is developed for secure storage and transmission of extensive medical data in medical images. The algorithm demonstrates high embedding capacity, imperceptibility, and resilience against statistical attacks in experimental evaluations. Incorporating this algorithm into the Internet of Medical Things (IoMT) sector enhances security measures for the storage and transmission of massive medical data, addressing the limitations of conventional 2D RDH algorithms in medical images.
Article
Chemistry, Analytical
Mohamed Hammad, Pawel Plawiak, Mohammed ElAffendi, Ahmed A. Abd El-Latif, Asmaa A. Abdel Latif
Summary: This study presents an enhanced deep learning approach, named Derma Care, for the accurate detection of eczema and psoriasis skin conditions. The proposed model addresses challenges faced by previous methods and achieves remarkable results in detecting multiple skin diseases simultaneously. The findings hold significant implications for dermatological diagnosis and early detection of skin diseases.
Article
Public, Environmental & Occupational Health
Haytham Al Ewaidat, Nagwan Abdel Samee, Jaya Prakash Allam, Kiran Kumar Patro, Pawel Plawiak, Noha F. Mahmoud, Mohamed Hammad
Summary: This article discusses the role and importance of medical imaging professionals in dealing with the COVID-19 pandemic in Jordan. Through a survey study, researchers found that medical imaging professionals have a high level of knowledge and perception about COVID-19, and they advise following the guidelines of CDC and WHO in healthcare settings.
HEALTH & SOCIAL CARE IN THE COMMUNITY
(2023)
