Aydin Akan - Academia.edu (original) (raw)

Papers by Aydin Akan

Social Science Research Network, 2022

Communications and Networking, 2010

2015 IEEE International Conference on Imaging Systems and Techniques (IST), 2015

Digital breast tomosynthesis (DBT) generates 3D images of breast by using 2D projections taken fr... more Digital breast tomosynthesis (DBT) generates 3D images of breast by using 2D projections taken from a limited view angle. Due to the ill-posed nature of the image reconstruction in DBT alternative image reconstruction methods have been introduced for better image quality. In this study, an efficient DBT image reconstruction algorithm has been introduced. The proposed method was formulated as combination of algebraic reconstruction technique (ART), total variation (TV) and non-local means (NLM). A real DBT data set and a commercially available 3D phantom were used for performance evaluations. Results of our study showed that ART+TV+NLM helped in obtaining better reconstructed images than those obtained by ART and ART+TV in terms of reduced background noise, out-of-plane artifacts while keeping the details well-preserved.

A lot of research has been done on sleep disorders from past to present. Sleep apnea, which we fr... more A lot of research has been done on sleep disorders from past to present. Sleep apnea, which we frequently encounter today, is one of the important sleep disorders that threaten human life. This situation that occurs during sleep also affects the daily life of the individual. Obstructive sleep apnea syndrome (OSAS) is a respiratory tract disorder with a prevalence of almost 4% in men and approximately 2% in women [1]. Snoring and OSA, which are among the breathing problems during sleep, are among the conditions caused by the insufficiency of breathing [2]. The aim of our study is to determine whether the person has OSA by analyzing electroencephalogram (EEG) signals. As we know, many physiological and biological activities occur during sleep. In order to observe these activities, we record the electrical activity that occur in our brain. Thanks to the EEG, we transform these activities into digital data. In this project, EEG signals recorded from 4 patients during sleep were processed on MATLAB. Sleep recordings of different sleep zones marked by the doctor are segmented. The data in the segments are divided into 3 headings as pre-apnea, moment of apnea and post-apnea. The data were processed with signal analysis methods such as empirical mode decomposition (EMD) and intrinsic mode functions (IMFs) were extracted. Attributes were obtained from IMFs again on MATLAB. These features are used for classification in advanced machine learning algorithms as pre-apnea and apnea moment as a set of 2 and as a set of 3 as pre-apnea, apnea moment and post-apnea. Using the method, we mentioned provides a practical and fast diagnostic process for patients and doctors in our project. In this project, which aims to accelerate the treatment and diagnosis process in order to support the health of patients, it is aimed to classify OSA by analyzing EEG signals. As a result of our project, the accuracy values of the 2-set are between 47.5% and 71.9%, and the accuracy values of the 3-set are between 33.8% - 63.1%.

Optics Communications, Apr 1, 2018

In this paper, a new spectrally efficient space modulation technique, which is called double quad... more In this paper, a new spectrally efficient space modulation technique, which is called double quadrature spatial intensity modulation (DQSIM), is proposed for multiple-input multiple-output (MIMO) visible light communication (VLC) systems. Sub-carrier intensity modulation (SCM), which ensures the use of in-phase/quadrature (I/Q) signals in intensity modulation direct detection (IM/DD) systems, is used as a digital modulation scheme. In RF, quadrature spatial modulation (QSM) transmits the I/Q signals through single or multiple antennas selected independently from each other. Furthermore, the orthogonality between I and Q components is provided for the half period of sinusoids. DQSIM utilizes these two features and transmits four fold more bits than spatial modulation (SM) via spatial constellation. SCM uses twofold bandwidth compared to on-off keying (OOK), while DQSIM uses three fold. DQSIM outperforms benchmark modulation schemes, which are SCM-SM and pulse amplitude modulation spatial modulation (PAM-SM), at the bit error rate (BER) value of 10-4. Furthermore DQSIM performance has increased with the increasing number of LEDs.

Social Science Research Network, 2022

Background: Electrocardiogram (ECG) is a method of recording the electrical activity of the heart... more Background: Electrocardiogram (ECG) is a method of recording the electrical activity of the heart and provides a diagnostic mean for heart-related diseases. An arrhythmia is any irregularity of heartbeat that causes an abnormality in one's heart rhythm. Early detection of arrhythmia has great importance to prevent many diseases. Manual analysis of ECG signal is not sufficient for quickly identifying arrhythmias that may cause sudden deaths. Hence, many studies have been presented to developed computer-aided diagnosis (CAD) systems to automatically identify arrhythmias. Methods: This paper proposes a novel deep learning approach to identify arrhythmias in ECG signals. The signals are obtained from MIT-BIH arrhythmia database and are categorized according to five arrhythmia types. The proposed approach identifies arrhythmia classes by using Convolutional Neural Network (CNN) architecture trained by twodimensional (2D) ECG beat images. CNN architecture is selected due to high image recognition performance. ECG signals are segmented into heartbeats, then each heartbeat is transformed into a 2D grayscale image. The heartbeat images are used as input for the CNN. Results: The proposed CNN model is compared to other common CNN architectures such as LeNet and ResNet-50 to evaluate the performance of our study. Overall, the proposed study achieved 99.7% test accuracy in the classification of five different ECG arrhythmias. Conclusions: Testing results demonstrate that CNN trained by ECG image representations provide outstanding classification performance of arrhythmic ECG signals and outperforms similar network architectures. Hence, the proposed approach provides a robust method for the classification of ECG arrhythmias.

Istanbul University - Journal of Electrical and Electronics Engineering, 2017

Journal of the Australian Ceramic Society, 2017

International Journal of Neural Systems, Sep 9, 2021

This research presents a new method for detecting obsessive–compulsive disorder (OCD) based on ti... more This research presents a new method for detecting obsessive–compulsive disorder (OCD) based on time–frequency analysis of multi-channel electroencephalogram (EEG) signals using the multi-variate synchrosqueezing transform (MSST). With the evolution of multi-channel sensor implementations, the employment of multi-channel techniques for the extraction of features arising from multi-channel dependency and mono-channel characteristics has become common. MSST has recently been proposed as a method for modeling the combined oscillatory mechanisms of multi-channel signals. It makes use of the concepts of instantaneous frequency (IF) and bandwidth. Electrophysiological data, like other nonstationary signals, necessitates both joint time–frequency analysis and independent time and frequency domain studies. The usefulness and effectiveness of a multi-variate, wavelet-based synchrosqueezing algorithm paired with a band extraction method are tested using electroencephalography data obtained from OCD patients and control groups in this research. The proposed methodology yields substantial results when analyzing differences between patient and control groups.

Elsevier eBooks, 2016

This chapter extends Part I by presenting additional advanced key principles underlying the use o... more This chapter extends Part I by presenting additional advanced key principles underlying the use of time-frequency (t,f) methods. The topic is covered in 11 focused sections.

International Journal of Neural Systems, May 26, 2021

Epilepsy is one of the most common brain disorders worldwide. The most frequently used clinical t... more Epilepsy is one of the most common brain disorders worldwide. The most frequently used clinical tool to detect epileptic events and monitor epilepsy patients is the EEG recordings. There have been proposed many computer-aided diagnosis systems using EEG signals for the detection and prediction of seizures. In this study, a novel method based on Fourier-based Synchrosqueezing Transform (SST), which is a high-resolution time-frequency (TF) representation, and Convolutional Neural Network (CNN) is proposed to detect and predict seizure segments. SST is based on the reassignment of signal components in the TF plane which provides highly localized TF energy distributions. Epileptic seizures cause sudden energy discharges which are well represented in the TF plane by using the SST method. The proposed SST-based CNN method is evaluated using the IKCU dataset we collected, and the publicly available CHB-MIT dataset. Experimental results demonstrate that the proposed approach yields high average segment-based seizure detection precision and accuracy rates for both datasets (IKCU: 98.99% PRE and 99.06% ACC; CHB-MIT: 99.81% PRE and 99.63% ACC). Additionally, SST-based CNN approach provides significantly higher segment-based seizure prediction performance with 98.54% PRE and 97.92% ACC than similar approaches presented in the literature using the CHB-MIT dataset.

Biomedical Signal Processing and Control, Mar 1, 2023

In the literature, several signal processing techniques have been used to diagnose epilepsy which... more In the literature, several signal processing techniques have been used to diagnose epilepsy which is a nervous system disease. However most of these techniques fail to analyse EEG signals which are dynamic and non-linear. In this study, an approach which utilizes a data-driven technique called Dynamic Mode Decomposition (DMD) that was originally developed to be used in fluid mechanics was proposed. Features that were belonged to EEG signals were calculated using DMD method and with the help of different classifiers, classification of the preseizure and seizure EEG signals was performed. Obtained results showed that the proposed method presented an alternative to approaches that are based on Empirical Mode Decomposition and its derivatives.

Alzheimer's Disease and Dementia are increasing diseases with the aging population. These dis... more Alzheimer's Disease and Dementia are increasing diseases with the aging population. These diseases cause memory loss, impaired attention, impaired problem-solving abilities. Serious games are designed to prevent or slow the progression of these diseases and to reduce the effects of diseases. In this study, a 2-dimensional maze game is designed as a serious game. The purpose of this study is to play the game by using EEG signals recording from the user.

Electrogastrography (EGG) is a non-invasive method used to assess the myoelectrical activity of t... more Electrogastrography (EGG) is a non-invasive method used to assess the myoelectrical activity of the stomach. It is frequently used by researchers and clinicians in diseases and conditions where gastric functions are impaired. Gastrointestinal system diseases such as functional dyspepsia and recurrent abdominal pain are very common in infants and children. In this study, previous studies done on EGG use in children were examined and in spite of the differences in recording time and methods, it's presented with examples that the use of EGG in children with other measurement methods is effective in disease diagnosis and treatment interventions.

Ultrasound imaging and Fine Needle Aspiration Biopsy, which used for diagnosis of thyroid cancer,... more Ultrasound imaging and Fine Needle Aspiration Biopsy, which used for diagnosis of thyroid cancer, don't ensure sufficient sensitivity and specificity. Due to this fact, many patients undergo unnecessary thyroid removal. In this study, it is aimed to find new features based on non-invasive Computerized Tomography. 52 nodular goiter patients who underwent thyroid removal surgery with suspicion of cancer were included in the study. Resected fresh thyroid tissues were imaged using Computerized Tomography (CT). In CT images, background and noise estimation as well as elimination were applied. Thyroid images were automatically segmented and modified histograms were calculated. Afterwards, features based on these histograms were calculated and it is proposed that these features can be used to discriminate malignant and benign thyroids. Moreover, as an indirect finding of histograms, it is concluded that thyroidal volume can be a useful criterion in determination of malignancy.

This paper presents a simple and fast approach to find a minimum sampling frequency for multi-ban... more This paper presents a simple and fast approach to find a minimum sampling frequency for multi-band signals. Instead of neighbor and boundary conditions, constraints on the sampling frequency were derived by using the geometric app­roach. Reformulation of the minimum sampling determination problem by using geometric approach enables to represent the problem as a basic inequality problem. Recursive algorithm was proposed to solve the constraints on the minimum sampling frequency. The proposed method was verified through numerical simulations in terms of the minimum sampling frequency and the computational efficiency by using 2-band and 3-band signals. Although the results illustrated the valid minimum sampling frequencies for the multi-band signals, due to the increase in the number of iterations, optimization approaches were recommended in the solution of the constraints on the minimum sampling frequency.

Yurume analizi ile ilgili gelismeler, bazi hastaliklarin ozelliklede norodejeneratif hastaliklari... more Yurume analizi ile ilgili gelismeler, bazi hastaliklarin ozelliklede norodejeneratif hastaliklarin teshisinde yurume analizinin yardimci bir faktor olarak yer almasini saglamistir. Ayagin altindaki kuvvete duyarli sensorler yardimi ile kontrol bireylerinden ve norodejeneratif hastalardan yurume isaretleri alinmistir. Bu isaretlere zaman-frekans analizi uygulanarak hastalikli isaretlerin normal isaretlerden farklari ortaya cikarilmistir. Daha sonra Yapay Sinir Aglari (YSA) ile yapilan siniflandirmada patolojik isaretler %82 dogrulukla tespit edilmistir.

Social Science Research Network, 2022

Communications and Networking, 2010

2015 IEEE International Conference on Imaging Systems and Techniques (IST), 2015

Digital breast tomosynthesis (DBT) generates 3D images of breast by using 2D projections taken fr... more Digital breast tomosynthesis (DBT) generates 3D images of breast by using 2D projections taken from a limited view angle. Due to the ill-posed nature of the image reconstruction in DBT alternative image reconstruction methods have been introduced for better image quality. In this study, an efficient DBT image reconstruction algorithm has been introduced. The proposed method was formulated as combination of algebraic reconstruction technique (ART), total variation (TV) and non-local means (NLM). A real DBT data set and a commercially available 3D phantom were used for performance evaluations. Results of our study showed that ART+TV+NLM helped in obtaining better reconstructed images than those obtained by ART and ART+TV in terms of reduced background noise, out-of-plane artifacts while keeping the details well-preserved.

A lot of research has been done on sleep disorders from past to present. Sleep apnea, which we fr... more A lot of research has been done on sleep disorders from past to present. Sleep apnea, which we frequently encounter today, is one of the important sleep disorders that threaten human life. This situation that occurs during sleep also affects the daily life of the individual. Obstructive sleep apnea syndrome (OSAS) is a respiratory tract disorder with a prevalence of almost 4% in men and approximately 2% in women [1]. Snoring and OSA, which are among the breathing problems during sleep, are among the conditions caused by the insufficiency of breathing [2]. The aim of our study is to determine whether the person has OSA by analyzing electroencephalogram (EEG) signals. As we know, many physiological and biological activities occur during sleep. In order to observe these activities, we record the electrical activity that occur in our brain. Thanks to the EEG, we transform these activities into digital data. In this project, EEG signals recorded from 4 patients during sleep were processed on MATLAB. Sleep recordings of different sleep zones marked by the doctor are segmented. The data in the segments are divided into 3 headings as pre-apnea, moment of apnea and post-apnea. The data were processed with signal analysis methods such as empirical mode decomposition (EMD) and intrinsic mode functions (IMFs) were extracted. Attributes were obtained from IMFs again on MATLAB. These features are used for classification in advanced machine learning algorithms as pre-apnea and apnea moment as a set of 2 and as a set of 3 as pre-apnea, apnea moment and post-apnea. Using the method, we mentioned provides a practical and fast diagnostic process for patients and doctors in our project. In this project, which aims to accelerate the treatment and diagnosis process in order to support the health of patients, it is aimed to classify OSA by analyzing EEG signals. As a result of our project, the accuracy values of the 2-set are between 47.5% and 71.9%, and the accuracy values of the 3-set are between 33.8% - 63.1%.

Optics Communications, Apr 1, 2018

In this paper, a new spectrally efficient space modulation technique, which is called double quad... more In this paper, a new spectrally efficient space modulation technique, which is called double quadrature spatial intensity modulation (DQSIM), is proposed for multiple-input multiple-output (MIMO) visible light communication (VLC) systems. Sub-carrier intensity modulation (SCM), which ensures the use of in-phase/quadrature (I/Q) signals in intensity modulation direct detection (IM/DD) systems, is used as a digital modulation scheme. In RF, quadrature spatial modulation (QSM) transmits the I/Q signals through single or multiple antennas selected independently from each other. Furthermore, the orthogonality between I and Q components is provided for the half period of sinusoids. DQSIM utilizes these two features and transmits four fold more bits than spatial modulation (SM) via spatial constellation. SCM uses twofold bandwidth compared to on-off keying (OOK), while DQSIM uses three fold. DQSIM outperforms benchmark modulation schemes, which are SCM-SM and pulse amplitude modulation spatial modulation (PAM-SM), at the bit error rate (BER) value of 10-4. Furthermore DQSIM performance has increased with the increasing number of LEDs.

Social Science Research Network, 2022

Background: Electrocardiogram (ECG) is a method of recording the electrical activity of the heart... more Background: Electrocardiogram (ECG) is a method of recording the electrical activity of the heart and provides a diagnostic mean for heart-related diseases. An arrhythmia is any irregularity of heartbeat that causes an abnormality in one's heart rhythm. Early detection of arrhythmia has great importance to prevent many diseases. Manual analysis of ECG signal is not sufficient for quickly identifying arrhythmias that may cause sudden deaths. Hence, many studies have been presented to developed computer-aided diagnosis (CAD) systems to automatically identify arrhythmias. Methods: This paper proposes a novel deep learning approach to identify arrhythmias in ECG signals. The signals are obtained from MIT-BIH arrhythmia database and are categorized according to five arrhythmia types. The proposed approach identifies arrhythmia classes by using Convolutional Neural Network (CNN) architecture trained by twodimensional (2D) ECG beat images. CNN architecture is selected due to high image recognition performance. ECG signals are segmented into heartbeats, then each heartbeat is transformed into a 2D grayscale image. The heartbeat images are used as input for the CNN. Results: The proposed CNN model is compared to other common CNN architectures such as LeNet and ResNet-50 to evaluate the performance of our study. Overall, the proposed study achieved 99.7% test accuracy in the classification of five different ECG arrhythmias. Conclusions: Testing results demonstrate that CNN trained by ECG image representations provide outstanding classification performance of arrhythmic ECG signals and outperforms similar network architectures. Hence, the proposed approach provides a robust method for the classification of ECG arrhythmias.

Istanbul University - Journal of Electrical and Electronics Engineering, 2017

Journal of the Australian Ceramic Society, 2017

International Journal of Neural Systems, Sep 9, 2021

This research presents a new method for detecting obsessive–compulsive disorder (OCD) based on ti... more This research presents a new method for detecting obsessive–compulsive disorder (OCD) based on time–frequency analysis of multi-channel electroencephalogram (EEG) signals using the multi-variate synchrosqueezing transform (MSST). With the evolution of multi-channel sensor implementations, the employment of multi-channel techniques for the extraction of features arising from multi-channel dependency and mono-channel characteristics has become common. MSST has recently been proposed as a method for modeling the combined oscillatory mechanisms of multi-channel signals. It makes use of the concepts of instantaneous frequency (IF) and bandwidth. Electrophysiological data, like other nonstationary signals, necessitates both joint time–frequency analysis and independent time and frequency domain studies. The usefulness and effectiveness of a multi-variate, wavelet-based synchrosqueezing algorithm paired with a band extraction method are tested using electroencephalography data obtained from OCD patients and control groups in this research. The proposed methodology yields substantial results when analyzing differences between patient and control groups.

Elsevier eBooks, 2016

This chapter extends Part I by presenting additional advanced key principles underlying the use o... more This chapter extends Part I by presenting additional advanced key principles underlying the use of time-frequency (t,f) methods. The topic is covered in 11 focused sections.

International Journal of Neural Systems, May 26, 2021

Epilepsy is one of the most common brain disorders worldwide. The most frequently used clinical t... more Epilepsy is one of the most common brain disorders worldwide. The most frequently used clinical tool to detect epileptic events and monitor epilepsy patients is the EEG recordings. There have been proposed many computer-aided diagnosis systems using EEG signals for the detection and prediction of seizures. In this study, a novel method based on Fourier-based Synchrosqueezing Transform (SST), which is a high-resolution time-frequency (TF) representation, and Convolutional Neural Network (CNN) is proposed to detect and predict seizure segments. SST is based on the reassignment of signal components in the TF plane which provides highly localized TF energy distributions. Epileptic seizures cause sudden energy discharges which are well represented in the TF plane by using the SST method. The proposed SST-based CNN method is evaluated using the IKCU dataset we collected, and the publicly available CHB-MIT dataset. Experimental results demonstrate that the proposed approach yields high average segment-based seizure detection precision and accuracy rates for both datasets (IKCU: 98.99% PRE and 99.06% ACC; CHB-MIT: 99.81% PRE and 99.63% ACC). Additionally, SST-based CNN approach provides significantly higher segment-based seizure prediction performance with 98.54% PRE and 97.92% ACC than similar approaches presented in the literature using the CHB-MIT dataset.

Biomedical Signal Processing and Control, Mar 1, 2023

In the literature, several signal processing techniques have been used to diagnose epilepsy which... more In the literature, several signal processing techniques have been used to diagnose epilepsy which is a nervous system disease. However most of these techniques fail to analyse EEG signals which are dynamic and non-linear. In this study, an approach which utilizes a data-driven technique called Dynamic Mode Decomposition (DMD) that was originally developed to be used in fluid mechanics was proposed. Features that were belonged to EEG signals were calculated using DMD method and with the help of different classifiers, classification of the preseizure and seizure EEG signals was performed. Obtained results showed that the proposed method presented an alternative to approaches that are based on Empirical Mode Decomposition and its derivatives.

Alzheimer's Disease and Dementia are increasing diseases with the aging population. These dis... more Alzheimer's Disease and Dementia are increasing diseases with the aging population. These diseases cause memory loss, impaired attention, impaired problem-solving abilities. Serious games are designed to prevent or slow the progression of these diseases and to reduce the effects of diseases. In this study, a 2-dimensional maze game is designed as a serious game. The purpose of this study is to play the game by using EEG signals recording from the user.

Electrogastrography (EGG) is a non-invasive method used to assess the myoelectrical activity of t... more Electrogastrography (EGG) is a non-invasive method used to assess the myoelectrical activity of the stomach. It is frequently used by researchers and clinicians in diseases and conditions where gastric functions are impaired. Gastrointestinal system diseases such as functional dyspepsia and recurrent abdominal pain are very common in infants and children. In this study, previous studies done on EGG use in children were examined and in spite of the differences in recording time and methods, it's presented with examples that the use of EGG in children with other measurement methods is effective in disease diagnosis and treatment interventions.

Ultrasound imaging and Fine Needle Aspiration Biopsy, which used for diagnosis of thyroid cancer,... more Ultrasound imaging and Fine Needle Aspiration Biopsy, which used for diagnosis of thyroid cancer, don't ensure sufficient sensitivity and specificity. Due to this fact, many patients undergo unnecessary thyroid removal. In this study, it is aimed to find new features based on non-invasive Computerized Tomography. 52 nodular goiter patients who underwent thyroid removal surgery with suspicion of cancer were included in the study. Resected fresh thyroid tissues were imaged using Computerized Tomography (CT). In CT images, background and noise estimation as well as elimination were applied. Thyroid images were automatically segmented and modified histograms were calculated. Afterwards, features based on these histograms were calculated and it is proposed that these features can be used to discriminate malignant and benign thyroids. Moreover, as an indirect finding of histograms, it is concluded that thyroidal volume can be a useful criterion in determination of malignancy.

This paper presents a simple and fast approach to find a minimum sampling frequency for multi-ban... more This paper presents a simple and fast approach to find a minimum sampling frequency for multi-band signals. Instead of neighbor and boundary conditions, constraints on the sampling frequency were derived by using the geometric app­roach. Reformulation of the minimum sampling determination problem by using geometric approach enables to represent the problem as a basic inequality problem. Recursive algorithm was proposed to solve the constraints on the minimum sampling frequency. The proposed method was verified through numerical simulations in terms of the minimum sampling frequency and the computational efficiency by using 2-band and 3-band signals. Although the results illustrated the valid minimum sampling frequencies for the multi-band signals, due to the increase in the number of iterations, optimization approaches were recommended in the solution of the constraints on the minimum sampling frequency.

Yurume analizi ile ilgili gelismeler, bazi hastaliklarin ozelliklede norodejeneratif hastaliklari... more Yurume analizi ile ilgili gelismeler, bazi hastaliklarin ozelliklede norodejeneratif hastaliklarin teshisinde yurume analizinin yardimci bir faktor olarak yer almasini saglamistir. Ayagin altindaki kuvvete duyarli sensorler yardimi ile kontrol bireylerinden ve norodejeneratif hastalardan yurume isaretleri alinmistir. Bu isaretlere zaman-frekans analizi uygulanarak hastalikli isaretlerin normal isaretlerden farklari ortaya cikarilmistir. Daha sonra Yapay Sinir Aglari (YSA) ile yapilan siniflandirmada patolojik isaretler %82 dogrulukla tespit edilmistir.