Article
Engineering, Multidisciplinary
Claudia Barile, Caterina Casavola, Giovanni Pappalettera, Vimalathithan Paramsamy Kannan
Summary: Signal-based acoustic emission data were analyzed in this research to identify damage modes in CFRP composites. Novel methodologies were introduced, and the 'dmey' wavelet was chosen for damage process identification through WPT, which showed consistent results with shifting in spectral density for characterizing damage modes.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2022)
Article
Mechanics
Jie Wang, Wei Zhou, Xia-ying Ren, Ming-ming Su, Jia Liu
Summary: A real-time analytical approach for damage mode identification of carbon fiber reinforced polymer using machine learning and acoustic emission is proposed. Waveform features are extracted from acoustic emission signals using wavelet packet transform, and a waveform-based clustering model is constructed to reveal the relevance between acoustic emission signals and damage modes. Different types of composite laminates can be recognized by the developed softmax layer classifier.
COMPOSITE STRUCTURES
(2023)
Article
Computer Science, Artificial Intelligence
Cosimo Ieracitano, Francesco Carlo Morabito, Amir Hussain, Nadia Mammone
Summary: In this paper, a hybrid-domain deep learning approach is proposed to decode hand movement preparation phases from EEG recordings, achieving a significant performance improvement compared to temporal-only and time-frequency-only-based methods, with an average accuracy of 76.21 +/- 3.77%. By combining temporal and time-frequency information using two CNNs and a standard multi-layer perceptron, a better classification result is achieved.
INTERNATIONAL JOURNAL OF NEURAL SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Ibrahim Mustafa Mehedi, Masoud Ahmadipour, Zainal Salam, Hussein Mohammed Ridha, Hussein Bassi, Muhyaddin Jamal Hosin Rawa, Mohammad Ajour, Abdullah Abusorrah, Md. Pauzi Abdullah
Summary: This paper introduces a method to enhance the performance of the multiclass support vector machine classifier using modified cuckoo search. Wavelet packet transform is used to extract crucial features for the classifier. Through simulations, the proposed method achieves high classification accuracies under different signal-to-noise ratios and outperforms other heuristic classification methods.
APPLIED SOFT COMPUTING
(2021)
Article
Engineering, Electrical & Electronic
Andrei S. Maliuk, Zahoor Ahmad, Jong-Myon Kim
Summary: This paper proposes a framework aimed at improving the accuracy of bearing-fault diagnosis. The framework utilizes a hybrid feature-selection method based on Wrapper-WPT. It decomposes the vibration signal using Wavelet Packet Transform and extracts time and frequency domain features. The features are then selected using the Boruta algorithm, and a Subspace k-NN is used for bearing fault diagnosis. The proposed method shows higher classification performance compared to other state-of-the-art methods.
Article
Computer Science, Hardware & Architecture
Hosein Eskandari, Maryam Imani, Mohsen Parsa Moghaddam
Summary: This paper proposes a short-term load forecasting method using a combination of wavelet transform and bidirectional gated recurrent unit. The best wavelet basis is selected using the Shannon entropy cost function. The selected features are applied to the BGRU for load forecasting. A new time coding approach is also designed to model the time patterns in the load time series. The proposed BT-WPT-BGRU method outperforms other methods in terms of the MAPE metric on the ISONE dataset.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Engineering, Biomedical
Dong Wang, Zhian Liu, Yi Tao, Wenjing Chen, Badong Chen, Qiang Wang, Xiangguo Yan, Gang Wang
Summary: The study introduces a novel EEG source imaging method (WPESI) based on wavelet packet transform and subspace component selection, which improves the localization accuracy of brain activity sources and outperforms traditional methods in computer simulations and experiments. For epilepsy patients, the activity sources estimated by this method align with seizure onset zones.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2021)
Article
Engineering, Multidisciplinary
Nitin Burud, J. M. Chandra Kishen
Summary: This work delves into the spectral realm of acoustic emission waveforms, proposing the use of wavelet entropy to estimate spectral disorder. It demonstrates the potential dual application of wavelet entropy as a signal discriminator and damage index. The increase in statistical variance of wavelet entropy distribution with stress level indicates the presence of multi-sources and multi-mechanistic fracture processes.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2021)
Article
Chemistry, Analytical
Huibin Zhu, Zhangming He, Juhui Wei, Jiongqi Wang, Haiyin Zhou
Summary: This paper proposes a bearing fault diagnosis method based on feature fusion, which extracts the time-frequency features of bearing signals through Wavelet Packet Transform and constructs Multi-Weight Singular Value Decomposition to effectively diagnose bearings. The proposed method shows better fault diagnosis and feature extraction capabilities compared to traditional methods.
Article
Computer Science, Information Systems
Ilkka Suuronen, Antti Airola, Tapio Pahikkala, Mika Murtojarvi, Valtteri Kaasinen, Henry Railo
Summary: Early detection plays a crucial role in future neuroprotective treatments for Parkinson's disease (PD). Resting state electroencephalographic (EEG) recording has the potential to be a cost-effective method for detecting neurological disorders such as PD. This study explored the impact of the number and placement of electrodes on classifying PD patients and healthy controls using machine learning based on EEG sample entropy. The results indicate that a small subset of electrodes placed in different areas of the brain can achieve classification performance comparable to using a full set of electrodes.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Acoustics
Li Ding, Jianxin Peng, Xiaowen Zhang, Lijuan Song
Summary: This paper proposes an automatic snoring detection system based on the wavelet packet transform (WPT) and eXtreme Gradient Boosting (XGBoost) classifier, which recognizes snoring sounds using the enhanced episodes by the generalization subspace noise reduction algorithm. The feature selection technology based on correlation analysis is applied to select the most discriminative WPT features. Experimental results show that WPT is effective in analyzing snoring and non-snoring sounds.
ARCHIVES OF ACOUSTICS
(2023)
Article
Engineering, Multidisciplinary
Yu-xing Li, Shang-bin Jiao, Xiang Gao
Summary: The article introduces a novel signal feature extraction technology based on EWT and RDE, which effectively extracts complex features of signals and improves signal separability and stability.
DEFENCE TECHNOLOGY
(2021)
Article
Engineering, Biomedical
Bethany Gosala, Pappu Dindayal Kapgate, Priyanka Jain, Rameshwar Nath Chaurasia, Manjari Gupta
Summary: Applying AI in healthcare benefits from Bio-signal analysis, particularly the Wavelet Scattering Transform (WST) method, which outperforms traditional ML algorithms (such as logistic regression and support vector machine) in neurological disorder classification. Results indicate that continuous wavelet transform (CWT) and discrete wavelet transform (DWT) yield better feature extraction performance. Decision trees achieve the best results in terms of accuracy, sensitivity, specificity, and Kappa score.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Engineering, Electrical & Electronic
Gholamreza Memarzadeh, Farshid Keynia
Summary: This paper presents a new hybrid forecast model for short-term electricity load and price prediction. By using wavelet transform, feature selection, and deep learning algorithm, the accuracy of predictions has been improved and successfully validated on actual data from multiple electricity markets.
ELECTRIC POWER SYSTEMS RESEARCH
(2021)
Review
Engineering, Biomedical
Syarifah Noor Syakiylla Sayed Daud, Rubita Sudirman
Summary: This review article comprehensively describes the application of the wavelet method in denoising the EEG signal based on recent research. It provides an overview of the basic theory and characteristics of EEG and the wavelet transform method, describes commonly applied wavelet-based methods for EEG dataset denoising, reviews a considerable number of the latest published EEG research works with wavelet applications, discusses challenges in current EEG-based wavelet method research, and recommends alternative solutions to mitigate the issues.
ANNALS OF BIOMEDICAL ENGINEERING
(2022)
Article
Engineering, Mechanical
Ersen Arslan, Caglar Uyulan
Summary: This study implements a system modeling based data analysis approach to evaluate the performance and stability of an e-scooter and rider system while driving over a curb. By solving over 20,000 different cases and using statistical analysis and machine learning techniques, the influence of various parameters on the system's output regarding stability and safety are reported and shared for further analysis by the e-scooter community.
PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART D-JOURNAL OF AUTOMOBILE ENGINEERING
(2023)
Article
Thermodynamics
Ayse Sagiroglu, Ilke Karagoz, Oznur Ozge Ozcan, Turker Tekin Erguzel, Mesut Karahan, Nevzat Tarhan
Summary: In this study, heat transfer in tumor treatment was simulated using radiofrequency and bioheat transfer modules, with a low error rate. A 2-D microwave antenna model was designed for hyperthermia treatment, and parameters for glioblastoma multiforme in brain tissue were determined.
NUMERICAL HEAT TRANSFER PART A-APPLICATIONS
(2023)
Article
Clinical Neurology
Caglar Uyulan, Turker Tekin Erguzel, Omer Turk, Shams Farhad, Baris Metin, Nevzat Tarhan
Summary: Automatic detection of ADHD based on fMRI using Deep Learning addresses the curse of-dimensionality problem and provides a robust solution. A transfer learning approach using ResNet-50 CNN achieved a classification accuracy of 93.45%, and Class Activation Map analysis revealed differences in brain regions between ADHD and healthy children.
CLINICAL EEG AND NEUROSCIENCE
(2023)
Article
Computer Science, Information Systems
Ilkim Ecem Emre, Cigdem Erol, Cumhur Tas, Nevzat Tarhan
Summary: The study analyzed EEG data of psychiatric patients using machine learning methods and found that differentiating between disease groups with high accuracy is feasible, especially excelling in recognizing diseases like ADHD, depression, and schizophrenia.
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS
(2023)
Article
Immunology
Gamze Gulden, Berranur Sert, Tarik Teymur, Yasin Ay, Nulifer Neslihan Tiryaki, Abhinava K. Mishra, Ercument Ovali, Nevzat Tarhan, Cihan Tastan
Summary: The development of genetic modification techniques has opened up a new era in cancer treatment by increasing the effectiveness and stability of CAR-T cell therapy. This study explores the use of Phytohemagglutinin (PHA) to increase the population of T cell memory phenotype in CAR-T cells, leading to long-term and efficient production of CAR-T cells.
Article
Computer Science, Interdisciplinary Applications
Zozan Guleken, Pawel Jakubczyk, Wieslaw Paja, Krzysztof Pancerz, Agnieszka Wosiak, Ilhan Yaylim, Guldal Inal Gultekin, Nevzat Tarhan, Mehmet Tolgahan Hakan, Dilara Sonmez, Devrim Saribal, Soykan Arikan, Joanna Depciuch
Summary: Gastric carcinoma ranks fifth in terms of incidence and third in terms of mortality globally. There is currently no accurate blood test for diagnosing gastric cancer. Raman spectroscopy was used to evaluate serum tumor marker levels and a prediction model was developed using machine learning techniques. The results suggest that Raman spectroscopy can be used to distinguish between gastric cancer patients and healthy individuals.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Clinical Neurology
Baris Metin, Shams Farhad, Turker Erguzel, Elvan Ciftci, Nevzat Tarhan
Summary: Diagnosing bipolar disorder can be challenging and delayed in clinical practice. This study investigates the use of both structural brain changes and neuropsychiatric tests together in an artificial neural network model for accurate classification of bipolar disorder.
JOURNAL OF NEURAL TRANSMISSION
(2023)
Article
Spectroscopy
Joanna Depciuch, Wieslaw Paja, Krzysztof Pancerz, Ozgur Uzun, Huri Bulut, Nevzat Tarhan, Zozan Guleken
Summary: Oocytes are supported by follicular fluid, and oxidative stress in the follicular fluid can affect oocyte development and embryo quality. This study used Raman spectroscopy combined with machine learning techniques to identify and quantify follicular fluid in patients with unexplained infertility. The results suggest that Raman spectroscopy can detect changes in follicular fluid associated with infertility.
JOURNAL OF RAMAN SPECTROSCOPY
(2023)
Article
Chemistry, Multidisciplinary
Caglar Uyulan, David Mayor, Tony Steffert, Tim Watson, Duncan Banks
Summary: This paper presents the results of using deep learning algorithms on EEG data to investigate the effects on the brain of different frequencies of TEAS. The study found that the greatest differences in EEG from baseline occurred during TEAS at 80 pulses per second (pps) or 160 pps, while the smallest differences occurred during 2.5 pps or 10 pps stimulation.
APPLIED SCIENCES-BASEL
(2023)
Article
Chemistry, Applied
Huri Bulut, Nevzat Tarhan, Melek Buyuk, Kursat Rahmi Serin, Engin Ulukaya, Joanna Depciuch, Magdalena Parlinska-Wojtan, Zozan Guleken
Summary: This study used FTIR spectroscopy to detect chemical changes in the serum of CCA patients and healthy individuals. The results showed decreased TAS and increased TOS, OSI, and total protein levels in CCA patients. The FTIR spectra were able to distinguish between CCA and controls, and changes in lipids and functional groups could be used to predict CCA.
BIOINTERFACE RESEARCH IN APPLIED CHEMISTRY
(2023)
Article
Engineering, Mechanical
Caglar Uyulan
Summary: This paper proposes a robust adaptive solution for stabilizing and tracking direct-drive flexible robot arms under parameter, model uncertainties, and external disturbances. The control approach considers the actuator dynamics, compensates unwanted torque ripples, and employs robust-adaptive input-output linearization-based control law combining chattering-free sliding mode control to achieve high-speed, precise position control while avoiding the excitation of unmodeled dynamics.