Mehmet Fatih Karakaş
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mehmet Fatih Karakaş has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Fatma Latifoğlu
Department of Biomedical Engineering, Erciyes University, Kayseri 38039, Türkiye
3 shared publications Esra Demirci
Tip Fakltesi ocuk Ergen ve Psikiyatri Anabilim Dali, Erciyes niversitesi, Kayseri, Trkiye
1 shared publicationFatma Latifoğlu
Biyomedikal Mhendislii Blm
1 shared publication
Latest Publications
- 20 April 2026
- Published by Eskisehir Osmangazi Universitesi Muhendislik ve Mimarlik Fakultesi Dergisi in Eskişehir Osmangazi Üniversitesi Mühendislik ve Mimarlık Fakültesi Dergisi
- Vol. 34 (1) , 2156-2166
DECODING BRAIN SIGNALS FOR INNER SPEECH: EEG SIGNAL SUB-BAND DECOMPOSITION AND MACHINE LEARNING APPROACH
Decoding inner speech is a rapidly advancing field with significant implications for brain-computer interfaces (BCIs) and assistive technologies. In this study how sub-band decomposition, feature types, and classifier choice affect inner-speech classification were assessed. A comprehensive comparative framework is presented, employing various sub-band decomposition techniques including Tunable Q-Factor Wavelet Transform, Empirical Mode Decomposition, Variational Mode Decomposition, and Circulant Singular Spectrum Analysis. From the decomposed signals, 17 statistical, 4 entropy, and 5 frequency features were extracted individually and in hybrid combinations before reduction with Least Absolute Shrinkage and Selection Operator. Classification performance was evaluated with 10-fold cross-validation, using Support Vector Machine, Feedforward Neural Network, and K-Nearest Neighbor. Performance was summarized as the mean across 30 independent runs per condition over six metrics. A three-way repeated measures ANOVA with Greenhouse-Geisser correction showed significant main effects and interactions across all metrics (p < 0.001), with large effect sizes (η2p >= 0.14). The best overall configuration was EMD-Hybrid-SVM outperforming baseline models, yielding a group-level mean accuracy of 54.13%. These findings demonstrate the potential of sub-band decomposition and hybrid features for improving inner speech decoding and contribute to the development of more robust EEG-based BCIs.
- 20 April 2026
- Published by Eskisehir Osmangazi Universitesi Muhendislik ve Mimarlik Fakultesi Dergisi in Eskişehir Osmangazi Üniversitesi Mühendislik ve Mimarlık Fakültesi Dergisi
- Vol. 34 (1) , 2156-2166
DECODING BRAIN SIGNALS FOR INNER SPEECH: EEG SIGNAL SUB-BAND DECOMPOSITION AND MACHINE LEARNING APPROACH
Decoding inner speech is a rapidly advancing field with significant implications for brain-computer interfaces (BCIs) and assistive technologies. In this study how sub-band decomposition, feature types, and classifier choice affect inner-speech classification were assessed. A comprehensive comparative framework is presented, employing various sub-band decomposition techniques including Tunable Q-Factor Wavelet Transform, Empirical Mode Decomposition, Variational Mode Decomposition, and Circulant Singular Spectrum Analysis. From the decomposed signals, 17 statistical, 4 entropy, and 5 frequency features were extracted individually and in hybrid combinations before reduction with Least Absolute Shrinkage and Selection Operator. Classification performance was evaluated with 10-fold cross-validation, using Support Vector Machine, Feedforward Neural Network, and K-Nearest Neighbor. Performance was summarized as the mean across 30 independent runs per condition over six metrics. A three-way repeated measures ANOVA with Greenhouse-Geisser correction showed significant main effects and interactions across all metrics (p < 0.001), with large effect sizes (η2p >= 0.14). The best overall configuration was EMD-Hybrid-SVM outperforming baseline models, yielding a group-level mean accuracy of 54.13%. These findings demonstrate the potential of sub-band decomposition and hybrid features for improving inner speech decoding and contribute to the development of more robust EEG-based BCIs.
- Open Access
- 30 June 2025
- Published by XLESCIENCE in International Journal of Advances in Signal and Image Sciences
- Vol. 11 (1) , 153-168
METAHEURISTIC-DRIVEN MULTI OBJECTIVE FIR FILTER DESIGN FOR ECG DENOISING
Digital signal processing in biomedical applications requires advanced filtering techniques that can simultaneously optimize multiple performance criteria. This study introduces a novel multi-objective metaheuristic approach for Finite Impulse Response (FIR) filter design, addressing the complex challenges of signal processing optimization. The research employs advanced metaheuristic algorithms, including Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), Genetic Algorithm (GA), Black Widow Optimization (BWO), Chaos Game Optimization (CGO), Harmony Search (HS), Squirrel Search Algorithm (SSA), and Atomic Orbital Search (AOS) to optimize multiple filter design parameters jointly. In contrast to conventional approaches, the proposed multi-objective optimization strategy demonstrates superior performance in signal filtering. The proposed method considers the important design trade-offs at the same time, such as minimizing signal power after stopband frequency, reducing stopband ripple, maximizing stopband first lobe and stopband edge frequency attenuation, and reducing computational complexity. The method is employed on ElectroCardioGram (ECG) signals from the MIT-BIH open-access database. Performance comparison indicated that multi-objective metaheuristic filters achieved much better performance than conventional FIR filters. The optimized filters exhibited improved stopband attenuation, narrower transition bands, lower power leakage, and more efficient signal processing at a cut-off frequency of 100 Hz. Signal power measurements demonstrated significant improvements. While conventional FIR filters ranged from -37.2001 dB to -41.46778 dB, multi-objective metaheuristic filters reached -42.05318 dB to -44.69498 dB in terms of stopband power.
- 30 June 2025
- Published by XLESCIENCE in International Journal of Advances in Signal and Image Sciences
- Vol. 11 (1) , 153-168
Open AccessMETAHEURISTIC-DRIVEN MULTI OBJECTIVE FIR FILTER DESIGN FOR ECG DENOISING
Digital signal processing in biomedical applications requires advanced filtering techniques that can simultaneously optimize multiple performance criteria. This study introduces a novel multi-objective metaheuristic approach for Finite Impulse Response (FIR) filter design, addressing the complex challenges of signal processing optimization. The research employs advanced metaheuristic algorithms, including Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC), Genetic Algorithm (GA), Black Widow Optimization (BWO), Chaos Game Optimization (CGO), Harmony Search (HS), Squirrel Search Algorithm (SSA), and Atomic Orbital Search (AOS) to optimize multiple filter design parameters jointly. In contrast to conventional approaches, the proposed multi-objective optimization strategy demonstrates superior performance in signal filtering. The proposed method considers the important design trade-offs at the same time, such as minimizing signal power after stopband frequency, reducing stopband ripple, maximizing stopband first lobe and stopband edge frequency attenuation, and reducing computational complexity. The method is employed on ElectroCardioGram (ECG) signals from the MIT-BIH open-access database. Performance comparison indicated that multi-objective metaheuristic filters achieved much better performance than conventional FIR filters. The optimized filters exhibited improved stopband attenuation, narrower transition bands, lower power leakage, and more efficient signal processing at a cut-off frequency of 100 Hz. Signal power measurements demonstrated significant improvements. While conventional FIR filters ranged from -37.2001 dB to -41.46778 dB, multi-objective metaheuristic filters reached -42.05318 dB to -44.69498 dB in terms of stopband power.
- Open Access
- 31 October 2023
- Published by Orclever Science and Research Group in The European Journal of Research and Development
- Vol. 3 (3) , 14-30
Complex Exponential Modulated Quadrature Filter Bank Design Based on Chaos Game Optimization Algorithm
The design approach presented in this paper uses a new metaheuristic method Chaos Game Optimization (CGO) to design a 2-channel quadrature mirror filter (QMF) banks. For this study, finite impulse response (FIR) low-pass prototype filter coefficients are optimized to minimize the cost function. The cost function is formulated as the sum of four terms: the square sum of the errors between this estimated frequency response and the ideal filter response, passband ripple, stopband ripple, and stopband edge frequency error. These parameters are important components of FIR filter prototype design that are often used alone in many studies. To obtain a QMF bank prototype filter, prototype filter modulated with an appropriate complex exponential signal. The produced design results of the proposed method like performance parameters, frequency responses of prototype filters and QMF bank results are presented and compared with earlier reported results. The results show that the presented study has significant improvement in the performance parameters of QMF banks and prototype filter.
- 31 October 2023
- Published by Orclever Science and Research Group in The European Journal of Research and Development
- Vol. 3 (3) , 14-30
Open AccessComplex Exponential Modulated Quadrature Filter Bank Design Based on Chaos Game Optimization Algorithm
The design approach presented in this paper uses a new metaheuristic method Chaos Game Optimization (CGO) to design a 2-channel quadrature mirror filter (QMF) banks. For this study, finite impulse response (FIR) low-pass prototype filter coefficients are optimized to minimize the cost function. The cost function is formulated as the sum of four terms: the square sum of the errors between this estimated frequency response and the ideal filter response, passband ripple, stopband ripple, and stopband edge frequency error. These parameters are important components of FIR filter prototype design that are often used alone in many studies. To obtain a QMF bank prototype filter, prototype filter modulated with an appropriate complex exponential signal. The produced design results of the proposed method like performance parameters, frequency responses of prototype filters and QMF bank results are presented and compared with earlier reported results. The results show that the presented study has significant improvement in the performance parameters of QMF banks and prototype filter.
- Open Access
- 17 May 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (10) , 1769
Distinguishing Parkinson’s Disease with GLCM Features from the Hankelization of EEG Signals
This study proposes a novel method that uses electroencephalography (EEG) signals to classify Parkinson’s Disease (PD) and demographically matched healthy control groups. The method utilizes the reduced beta activity and amplitude decrease in EEG signals that are associated with PD. The study involved 61 PD patients and 61 demographically matched controls groups, and EEG signals were recorded in various conditions (eyes closed, eyes open, eyes both open and closed, on-drug, off-drug) from three publicly available EEG data sources (New Mexico, Iowa, and Turku). The preprocessed EEG signals were classified using features obtained from gray-level co-occurrence matrix (GLCM) features through the Hankelization of EEG signals. The performance of classifiers with these novel features was evaluated using extensive cross-validations (CV) and leave-one-out cross-validation (LOOCV) schemes. This method under 10 × 10 fold CV, the method was able to differentiate PD groups from healthy control groups using a support vector machine (SVM) with an accuracy of 92.4 ± 0.01, 85.7 ± 0.02, and 77.1 ± 0.06 for New Mexico, Iowa, and Turku datasets, respectively. After a head-to-head comparison with state-of-the-art methods, this study showed an increase in the classification of PD and controls.
- 17 May 2023
- Published by MDPI AG in Diagnostics
- Vol. 13 (10) , 1769
Open AccessDistinguishing Parkinson’s Disease with GLCM Features from the Hankelization of EEG Signals
This study proposes a novel method that uses electroencephalography (EEG) signals to classify Parkinson’s Disease (PD) and demographically matched healthy control groups. The method utilizes the reduced beta activity and amplitude decrease in EEG signals that are associated with PD. The study involved 61 PD patients and 61 demographically matched controls groups, and EEG signals were recorded in various conditions (eyes closed, eyes open, eyes both open and closed, on-drug, off-drug) from three publicly available EEG data sources (New Mexico, Iowa, and Turku). The preprocessed EEG signals were classified using features obtained from gray-level co-occurrence matrix (GLCM) features through the Hankelization of EEG signals. The performance of classifiers with these novel features was evaluated using extensive cross-validations (CV) and leave-one-out cross-validation (LOOCV) schemes. This method under 10 × 10 fold CV, the method was able to differentiate PD groups from healthy control groups using a support vector machine (SVM) with an accuracy of 92.4 ± 0.01, 85.7 ± 0.02, and 77.1 ± 0.06 for New Mexico, Iowa, and Turku datasets, respectively. After a head-to-head comparison with state-of-the-art methods, this study showed an increase in the classification of PD and controls.
- 1 October 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Assessment of anxiety disorder by pupil response
Specific phobia which is a state of excessive and constant fear although individuals know it is illogical. During phobia in individuals there is several changes. This study based on recorded pupil video during phobia and analysis of changes in pupil size using image processing and signal processing algorithms on this video. Feature extraction are made from the signals obtained from the pupil dimensions. These features were statistically analyzed to determine differences between groups. As a result of this analysis laid out attributes will be important in the diagnosis and monitoring of phobia.
- 1 October 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Assessment of anxiety disorder by pupil response
Specific phobia which is a state of excessive and constant fear although individuals know it is illogical. During phobia in individuals there is several changes. This study based on recorded pupil video during phobia and analysis of changes in pupil size using image processing and signal processing algorithms on this video. Feature extraction are made from the signals obtained from the pupil dimensions. These features were statistically analyzed to determine differences between groups. As a result of this analysis laid out attributes will be important in the diagnosis and monitoring of phobia.