Özal Yıldırım

The novel coronavirus 2019 (COVID-2019), which first appeared in Wuhan city of China in December 2019, spread rapidly around the world and became a pandemic. It has caused a devastating effect on both daily lives, public health, and the global economy. It is critical to detect the positive cases as early as possible so as to prevent the further spread of this epidemic and to quickly treat affected patients. The need for auxiliary diagnostic tools has increased as there are no accurate automated toolkits available. Recent findings obtained using radiology imaging techniques suggest that such images contain salient information about the COVID-19 virus. Application of advanced artificial intelligence (AI) techniques coupled with radiological imaging can be helpful for the accurate detection of this disease, and can also be assistive to overcome the problem of a lack of specialized physicians in remote villages. In this study, a new model for automatic COVID-19 detection using raw chest X-ray images is presented. The proposed model is developed to provide accurate diagnostics for binary classification (COVID vs. No-Findings) and multi-class classification (COVID vs. No-Findings vs. Pneumonia). Our model produced a classification accuracy of 98.08% for binary classes and 87.02% for multi-class cases. The DarkNet model was used in our study as a classifier for the you only look once (YOLO) real time object detection system. We implemented 17 convolutional layers and introduced different filtering on each layer. Our model (available at (https://github.com/muhammedtalo/COVID-19)) can be employed to assist radiologists in validating their initial screening, and can also be employed via cloud to immediately screen patients.

This article presents a new deep learning approach for cardiac arrhythmia (17 classes) detection based on long-duration electrocardiography (ECG) signal analysis. Cardiovascular disease prevention is one of the most important tasks of any health care system as about 50 million people are at risk of heart disease in the world. Although automatic analysis of ECG signal is very popular, current methods are not satisfactory. The goal of our research was to design a new method based on deep learning to efficiently and quickly classify cardiac arrhythmias. Described research are based on 1000 ECG signal fragments from the MIT - BIH Arrhythmia database for one lead (MLII) from 45 persons. Approach based on the analysis of 10-s ECG signal fragments (not a single QRS complex) is applied (on average, 13 times less classifications/analysis). A complete end-to-end structure was designed instead of the hand-crafted feature extraction and selection used in traditional methods. Our main contribution is to design a new 1D-Convolutional Neural Network model (1D-CNN). The proposed method is 1) efficient, 2) fast (real-time classification) 3) non-complex and 4) simple to use (combined feature extraction and selection, and classification in one stage). Deep 1D-CNN achieved a recognition overall accuracy of 17 cardiac arrhythmia disorders (classes) at a level of 91.33% and classification time per single sample of 0.015 s. Compared to the current research, our results are one of the best results to date, and our solution can be implemented in mobile devices and cloud computing.

Long-short term memory networks (LSTMs), which have recently emerged in sequential data analysis, are the most widely used type of recurrent neural networks (RNNs) architecture. Progress on the topic of deep learning includes successful adaptations of deep versions of these architectures. In this study, a new model for deep bidirectional LSTM network-based wavelet sequences called DBLSTM-WS was proposed for classifying electrocardiogram (ECG) signals. For this purpose, a new wavelet-based layer is implemented to generate ECG signal sequences. The ECG signals were decomposed into frequency sub-bands at different scales in this layer. These sub-bands are used as sequences for the input of LSTM networks. New network models that include unidirectional (ULSTM) and bidirectional (BLSTM) structures are designed for performance comparisons. Experimental studies have been performed for five different types of heartbeats obtained from the MIT-BIH arrhythmia database. These five types are Normal Sinus Rhythm (NSR), Ventricular Premature Contraction (VPC), Paced Beat (PB), Left Bundle Branch Block (LBBB), and Right Bundle Branch Block (RBBB). The results show that the DBLSTM-WS model gives a high recognition performance of 99.39%. It has been observed that the wavelet-based layer proposed in the study significantly improves the recognition performance of conventional networks. This proposed network structure is an important approach that can be applied to similar signal processing problems.

Magnetic resonance imaging (MRI) is the most common imaging technique used to detect abnormal brain tumors. Traditionally, MRI images are analyzed manually by radiologists to detect the abnormal conditions in the brain. Manual interpretation of huge volume of images is time consuming and difficult. Hence, computer-based detection helps in accurate and fast diagnosis. In this study, we proposed an approach that uses deep transfer learning to automatically classify normal and abnormal brain MR images. Convolutional neural network (CNN) based ResNet34 model is used as a deep learning model. We have used current deep learning techniques such as data augmentation, optimal learning rate finder and fine-tuning to train the model. The proposed model achieved 5-fold classification accuracy of 100% on 613 MR images. Our developed system is ready to test on huge database and can assist the radiologists in their daily screening of MR images.

Myocardial infarction (MI), commonly known as heart attack, causes irreversible damage to heart muscles and even leads to death. Rapid and accurate diagnosis of MI is critical to avoid death. Blood tests and electrocardiogram (ECG) signals are used to diagnose acute MI. However, for an increase in blood enzyme values, a certain time must pass after the attack. This time lag may delay MI diagnosis. Hence, ECG diagnosis is still very important. Manual ECG interpretation requires expertise and is prone to inter-observer variability. Therefore, computer aided diagnosis may be useful in automatic detection of MI on ECG. In this study, a deep learning model with an end-to-end structure on the standard 12-lead ECG signal for the diagnosis of MI is proposed. For this purpose, the most commonly used technique, convolutional neural network (CNN) is used. Our trained CNN model with the proposed architecture yielded impressive accuracy and sensitivity performance over 99.00% for MI diagnosis on all ECG lead signals. Thus, the proposed model has the potential to provide high performance on MI detection which can be used in wearable technologies and intensive care units.

For diagnosis of arrhythmic heart problems, electrocardiogram (ECG) signals should be recorded and monitored. The long-term signal records obtained are analyzed by expert cardiologists. Devices such as the Holter monitor have limited hardware capabilities. For improved diagnostic capacity, it would be helpful to detect arrhythmic signals automatically. In this study, a novel approach is presented as a candidate solution for these issues.A convolutional auto-encoder (CAE) based nonlinear compression structure is implemented to reduce the signal size of arrhythmic beats. Long-short term memory (LSTM) classifiers are employed to automatically recognize arrhythmias using ECG features, which are deeply coded with the CAE network.Based upon the coded ECG signals, both storage requirement and classification time were considerably reduced. In experimental studies conducted with the MIT-BIH arrhythmia database, ECG signals were compressed by an average 0.70% percentage root mean square difference (PRD) rate, and an accuracy of over 99.0% was observed.One of the significant contributions of this study is that the proposed approach can significantly reduce time duration when using LSTM networks for data analysis. Thus, a novel and effective approach was proposed for both ECG signal compression, and their high-performance automatic recognition, with very low computational cost.

Advances in artificial intelligence technologies have made it possible to obtain more accurate and reliable results using digital images. Due to the advances in digital histopathological images obtained using whole slide image (WSI) scanners, automated analysis of digital images by computer support systems has become interesting. In particular, deep learning architectures, are one of the preferred approaches in the analysis of digital histopathology images. The deeper networks trained on large amounts of image data are adapted for different tasks using transfer learning technique. In this study, automated detection of invasive ductal carcinoma (IDC), which is the most common subtype of breast cancers, is proposed using deep transfer learning technique. We have used deep learning pre-trained models, ResNet-50 and DenseNet-161 for the IDC detection task. The public histopathology dataset containing 277,524 image patches were used in our experimental studies. As a result of training on the last layers of pre-trained deep networks, DenseNet-161 model has yielded F-sore of 92.38% and balanced accuracy value of 91.57%. Similarly, we have obtained F-score of 94.11% and balanced accuracy value of 90.96% using ResNet-50 architecture. In addition, our developed model is validated using the publicly available BreakHis breast cancer dataset and obtained promising results in classifying magnification independent histopathology images into benign and malignant classes. Our developed system obtained the highest classification performance as compared to the state-of-art techniques and is ready to be tested with more diverse huge databases.

The brain disorders may cause loss of some critical functions such as thinking, speech, and movement. So, the early detection of brain diseases may help to get the timely best treatment. One of the conventional methods used to diagnose these disorders is the magnetic resonance imaging (MRI) technique. Manual diagnosis of brain abnormalities is time-consuming and difficult to perceive the minute changes in the MRI images, especially in the early stages of abnormalities. Proper selection of the features and classifiers to obtain the highest performance is a challenging task. Hence, deep learning models have been widely used for medical image analysis over the past few years. In this study, we have employed the AlexNet, Vgg-16, ResNet-18, ResNet-34, and ResNet-50 pre-trained models to automatically classify MR images in to normal, cerebrovascular, neoplastic, degenerative, and inflammatory diseases classes. We have also compared their classification performance with pre-trained models, which are the state-of-art architectures. We have obtained the best classification accuracy of 95.23% ± 0.6 with the ResNet-50 model among the five pre-trained models. Our model is ready to be tested with huge MRI images of brain abnormalities. The outcome of the model will also help the clinicians to validate their findings after manual reading of the MRI images.

Deep learning models have become a popular mode to classify electrocardiogram (ECG) data. Investigators have used a variety of deep learning techniques for this application. Herein, a detailed examination of deep learning methods for ECG arrhythmia detection is provided. Approaches used by investigators are examined, and their contributions to the field are detailed. For this purpose, journal papers have been surveyed according to the methods used. In addition, various deep learning models and experimental studies are described and discussed. A five-class ECG dataset containing 100,022 beats was then utilized for further analysis of deep learning techniques. The constructed models were examined with this dataset, and results are presented. This study therefore provides information concerning deep learning approaches used for arrhythmia classification, and suggestions for further research in this area.

Face recognition is of great importance to real world applications such as video surveillance, human machine interaction and security systems. As compared to traditional machine learning approaches, deep learning based methods have shown better performances in terms of accuracy and speed of processing in image recognition. This paper proposes a modified Convolutional Neural Network (CNN) architecture by adding two normalization operations to two of the layers. The normalization operation which is batch normalization provided acceleration of the network. CNN architecture was employed to extract distinctive face features and Softmax classifier was used to classify faces in the fully connected layer of CNN. In the experiment part, Georgia Tech Database showed that the proposed approach has improved the face recognition performance with better recognition results.

Sleep disorder is a symptom of many neurological diseases that may significantly affect the quality of daily life. Traditional methods are time-consuming and involve the manual scoring of polysomnogram (PSG) signals obtained in a laboratory environment. However, the automated monitoring of sleep stages can help detect neurological disorders accurately as well. In this study, a flexible deep learning model is proposed using raw PSG signals. A one-dimensional convolutional neural network (1D-CNN) is developed using electroencephalogram (EEG) and electrooculogram (EOG) signals for the classification of sleep stages. The performance of the system is evaluated using two public databases (sleep-edf and sleep-edfx). The developed model yielded the highest accuracies of 98.06%, 94.64%, 92.36%, 91.22%, and 91.00% for two to six sleep classes, respectively, using the sleep-edf database. Further, the proposed model obtained the highest accuracies of 97.62%, 94.34%, 92.33%, 90.98%, and 89.54%, respectively for the same two to six sleep classes using the sleep-edfx dataset. The developed deep learning model is ready for clinical usage, and can be tested with big PSG data.

Advances in information technology have facilitated the retrieval and processing of biomedical data. Especially with wearable technologies and mobile platforms, we are able to follow our healthcare data, such as electrocardiograms (ECG), in real time. However, the hardware resources of these technologies are limited. For this reason, the optimal storage and safe transmission of the personal health data is critical. This study proposes a new deep convolutional autoencoder (CAE) model for compressing ECG signals. In this paper, a deep network structure of 27 layers consisting of encoder and decoder parts is designed. In the encoder section of this model, the signals are reduced to low-dimensional vectors; and in the decoder section, the signals are reconstructed. The deep learning approach provides the representations of the low and high levels of signals in the hidden layers of the model. Hence, the original signal can be reconstructed with minimal loss. Very different from traditional linear transformation methods, a deep compression approach implies that it can learn to use different ECG records automatically. The performance was evaluated on an experimental data set comprising 4800 ECG fragments from 48 unique clinical patients. The compression rate (CR) of the proposed model was 32.25, and the average PRD value was 2.73%. These favourable observation suggest that our deep model can allow secure data transfer in a low-dimensional form to remote medical centers. We present an effective compression approach that can potentially be used in wearable devices, e-health applications, telemetry and Holter systems.

Abstract In this paper, a new optimal feature selection based power quality event recognition system is proposed for the classification of power quality events. While Apriori algorithm is capable of processing categorical data, an effective feature vector, which represents distinctive features of digital power quality event data, has been obtained by means of the proposed k -means based Apriori algorithm feature selection approach. The proposed k -means based Apriori algorithm feature selection approach is presented with a power quality event recognition system. In the power quality event recognition system, normalization and segmentation processes have been applied to three-phase event voltage signals. Using 9-level multiresolution analysis, wavelet transform coefficients of the event signals have been obtained. By applying nine different feature extraction processes to these coefficients, a 90 dimensional feature vector belonging to three-phase event voltage signals has been extracted. Optimal feature vector has been obtained by applying the k -means based Apriori algorithm feature selection approach to the obtained feature vector, which has been applied as the last step to the input of the least squares support vector machine classifier and recognition performance results have been obtained. Real power quality event data have been used to evaluate the performance of the proposed feature selection approach and power quality event recognition system. According to the results, the proposed k -means based Apriori algorithm feature selection approach and power quality event recognition system are efficient, reliable and applicable and classify three-phase event types with a high degree of accuracy.

In this study, a deep-transfer learning approach is proposed for the automated diagnosis of diabetes mellitus (DM), using heart rate (HR) signals obtained from electrocardiogram (ECG) data. Recent progress in deep learning has contributed significantly to improvement in the quality of healthcare. In order for deep learning models to perform well, large datasets are required for training. However, a difficulty in the biomedical field is the lack of clinical data with expert annotation. A recent, commonly implemented technique to train deep learning models using small datasets is to transfer the weighting, developed from a large dataset, to the current model. This deep learning transfer strategy is generally employed for two-dimensional signals. Herein, the weighting of models pre-trained using two-dimensional large image data was applied to one-dimensional HR signals. The one-dimensional HR signals were then converted into frequency spectrum images, which were utilized for application to well-known pre-trained models, specifically: AlexNet, VggNet, ResNet, and DenseNet. The DenseNet pre-trained model yielded the highest classification average accuracy of 97.62%, and sensitivity of 100%, to detect DM subjects via HR signal recordings. In the future, we intend to further test this developed model by utilizing additional data along with cloud-based storage to diagnose DM via heart signal analysis.

In this paper, a new S-Transform and Extreme Learning Machine (ST–ELM)-based event recognition approach for the purpose of classifying power quality (PQ) event signals automatically has been proposed. In this approach, the distinctive features of the PQ event signals have been obtained with the S-Transform-based feature extraction. The feature vector obtained with feature extraction has been applied as input to the ELM classifier. Ten different classification procedures were determined within the framework of this study to assess the performance of the ELM classifier on PQ event data. Real PQ event data and synthetic PQ event data obtained from MATLAB/Simulink environment have been used in these procedures. Also, three different PQ event data sets, which are formed by adding noises of 20, 30 and 50 dB to the synthetic PQ event data respectively, have been used in order to assess the performance of the proposed approach on noisy conditions. According to the results of performance evaluations, the proposed ST–ELM-based PQ event recognition system has a very high performance of recognizing PQ event data. Besides, classification of noisy data showed that the proposed approach is robust at recognizing noisy data. The performance of the ST–ELM-based recognition system on PQ data shows that this approach has an effective recognition structure that can be used in real power systems.

Cardiac arrhythmia, which is an abnormal heart rhythm, is a common clinical problem in cardiology. Detection of arrhythmia on an extended duration electrocardiogram (ECG) is done based on initial algorithmic software screening, with final visual validation by cardiologists. It is a time consuming and subjective process. Therefore, fully automated computer-assisted detection systems with a high degree of accuracy have an essential role in this task. In this study, we proposed an effective deep neural network (DNN) model to detect different rhythm classes from a new ECG database. Our DNN model was designed for high performance on all ECG leads. The proposed model, which included both representation learning and sequence learning tasks, showed promising results on all 12-lead inputs. Convolutional layers and sub-sampling layers were used in the representation learning phase. The sequence learning part involved a long short-term memory (LSTM) unit after representation of learning layers. We performed two different class scenarios, including reduced rhythms (seven rhythm types) and merged rhythms (four rhythm types) according to the records from the database. Our trained DNN model achieved 92.24% and 96.13% accuracies for the reduced and merged rhythm classes, respectively. Recently, deep learning algorithms have been found to be useful because of their high performance. The main challenge is the scarcity of appropriate training and testing resources because model performance is dependent on the quality and quantity of case samples. In this study, we used a new public arrhythmia database comprising more than 10,000 records. We constructed an efficient DNN model for automated detection of arrhythmia using these records.

This paper offers an overview of essential concepts in deep learning, one of the state of the art approaches in machine learning, in terms of its history and current applications as a brief introduction to the subject. Deep learning has shown great successes in many domains such as handwriting recognition, image recognition, object detection etc. We revisited the concepts and mechanisms of typical deep learning algorithms such as Convolutional Neural Networks, Recurrent Neural Networks, Restricted Boltzmann Machine, and Autoencoders. We provided an intuition to deep learning that does not rely heavily on its deep math or theoretical constructs.

Kidney stones are a common complaint worldwide, causing many people to admit to emergency rooms with severe pain. Various imaging techniques are used for the diagnosis of kidney stone disease. Specialists are needed for the interpretation and full diagnosis of these images. Computer-aided diagnosis systems are the practical approaches that can be used as auxiliary tools to assist the clinicians in their diagnosis. In this study, an automated detection of kidney stone (having stone/not) using coronal computed tomography (CT) images is proposed with deep learning (DL) technique which has recently made significant progress in the field of artificial intelligence. A total of 1799 images were used by taking different cross-sectional CT images for each person. Our developed automated model showed an accuracy of 96.82% using CT images in detecting the kidney stones. We have observed that our model is able to detect accurately the kidney stones of even small size. Our developed DL model yielded superior results with a larger dataset of 433 subjects and is ready for clinical application. This study shows that recently popular DL methods can be employed to address other challenging problems in urology.

<h2>Abstract</h2><h3>Purpose</h3> Cardiovascular disease (CVD) is a leading cause of death globally. Electrocardiogram (ECG), which records the electrical activity of the heart, has been used for the diagnosis of CVD. The automated and robust detection of CVD from ECG signals plays a significant role for early and accurate clinical diagnosis. The purpose of this study is to provide automated detection of coronary artery disease (CAD) from ECG signals using capsule networks (CapsNet). <h3>Methods</h3> Deep learning-based approaches have become increasingly popular in computer aided diagnosis systems. Capsule networks are one of the new promising approaches in the field of deep learning. In this study, we used 1D version of CapsNet for the automated detection of coronary artery disease (CAD) on two second (95,300) and five second-long (38,120) ECG segments. These segments are obtained from 40 normal and 7 CAD subjects. In the experimental studies, 5-fold cross validation technique is employed to evaluate performance of the model. <h3>Results</h3> The proposed model, which is named as 1D-CADCapsNet, yielded a promising 5-fold diagnosis accuracy of 99.44% and 98.62% for two- and five-second ECG signal groups, respectively. We have obtained the highest performance results using 2 s ECG segment than the state-of-art studies reported in the literature. <h3>Conclusions</h3> 1D-CADCapsNet model automatically learns the pertinent representations from raw ECG data without using any hand-crafted technique and can be used as a fast and accurate diagnostic tool to help cardiologists.

Dental caries is the most frequent dental health issue in the general population. Dental caries can result in extreme pain or infections, lowering people's quality of life. Applying machine learning models to automatically identify dental caries can lead to earlier treatment. However, physicians frequently find the model results unsatisfactory due to a lack of explainability. Our study attempts to address this issue with an explainable deep learning model for detecting dental caries. We tested three prominent pre-trained models, EfficientNet-B0, DenseNet-121, and ResNet-50, to determine which is best for the caries detection task. These models take panoramic images as the input, producing a caries-non-caries classification result and a heat map, which visualizes areas of interest on the tooth. The model performance was evaluated using whole panoramic images of 562 subjects. All three models produced remarkably similar results. However, the ResNet-50 model exhibited a slightly better performance when compared to EfficientNet-B0 and DenseNet-121. This model obtained an accuracy of 92.00%, a sensitivity of 87.33%, and an F1-score of 91.61%. Visual inspection showed us that the heat maps were also located in the areas with caries. The proposed explainable deep learning model diagnosed dental caries with high accuracy and reliability. The heat maps help to explain the classification results by indicating a region of suspected caries on the teeth. Dentists could use these heat maps to validate the classification results and reduce misclassification.

Kidney stone disease is a serious public health concern that is getting worse with changes in diet, obesity, medical conditions, certain supplements etc. A kidney stone also called a renal calculus, is a hard buildup of urine minerals that form in the kidneys. Computed tomography (CT) is one of the imaging models used to identify kidney stones by clinical experts. Due to the low resolution of these images, sometimes detecting kidney stones is tedious with the naked eye, which may lead to false alarms. In this work, a computer-based diagnosis system with a deep learning technique has been developed as a practical solution to aid clinicians in their diagnosis. The traditional convolutional neural network (CNN)-based deep learning technology can detect stones in the kidney. Still, it suffers from the performance and standard implementation of the convolution operations in convolution layers. A Kronecker product-based convolution technique is incorporated in the proposed deep learning architecture to reduce the redundancy in feature maps without convolution overlapping. Our proposed method helps to make the network more effective by extracting abstract and in-depth features from the input images. The publicly available GitHub kidney stone CT scans are utilized to develop the proposed architecture. Our automated model detected kidney stones with an accuracy of 98.56% utilizing CT images. Our system is more effective than the most recent and cutting-edge techniques developed for identifying kidney stones of any size, including the smallest ones.

The diagnosis of urinary tract infections and kidney diseases using urine microscopy images has gained significant attention of medical community in recent years. These images are usually created by physicians’ own rule of thumb manually. However, this manual urine sediment analysis is usually labor-intensive and time-consuming. In addition, even when physicians carefully examine an image, an erroneous cell recognition may occur due to some optical illusions. In order to achieve cell recognition in low-resolution urine microscopy images with a higher level of accuracy, a new super resolution Faster Region-based Convolutional Neural Network (Faster R-CNN) method is proposed. It aims to increase resolution in low-resolution urine microscopy images using self-similarity based single image super resolution which was used during the pre-processing. De-noising based Wiener filter and Discrete Wavelet Transform (DWT) are used to de-noise high resolution images, respectively, to increase the level of accuracy for image recognition. Finally, for the feature extraction and classification stages, AlexNet, VGFG16 and VGG19 based Faster R-CNN models are used for the recognition and detection of multi-class cells. The model yielded accuracy rates are 98.6%, 96.4% and 96.2% respectively.

Atrial fibrillation (AF) is a common arrhythmia that can lead to stroke, heart failure, and premature death. Manual screening of AF on electrocardiography (ECG) is time-consuming and prone to errors. To overcome these limitations, computer-aided diagnosis systems are developed using artificial intelligence techniques for automated detection of AF. Various machine learning and deep learning (DL) techniques have been developed for the automated detection of AF. In this review, we focused on the automated AF detection models developed using DL techniques. Twenty-four relevant articles published in international journals were reviewed. DL models based on deep neural network, convolutional neural network (CNN), recurrent neural network, long short-term memory, and hybrid structures were discussed. Our analysis showed that the majority of the studies used CNN models, which yielded the highest detection performance using ECG and heart rate variability signals. Details of the ECG databases used in the studies, performance metrics of the various models deployed, associated advantages and limitations, as well as proposed future work were summarized and discussed. This review paper serves as a useful resource for the researchers interested in developing innovative computer-assisted ECG-based DL approaches for AF detection.

In this paper, a real time embedded intelligent recognition system is proposed for diagnosing the power quality problems. The intelligent recognition system has a structure capable of classifying and detecting the power quality problems in real time. Hardware applications of the wavelet transform and decision tree classifier are realized inside the recognition system using with field programmable gate array (FPGA) device. These methods operating as embedded in the FPGA environment provide real-time diagnosis of power quality problems. A new approach of this recognition system is its capability to simultaneously detect a power quality event in the power systems and power quality disturbances occurring on each phase following the event. In this paper, 2 different recognition systems that online and offline are presented. The online recognition system operates in the fields with signal processing and classification structures embedded. In the offline system, the distinctive features of the event signals obtained from the online system are used and these signals are classified in the computer environment by means of the least square support vector machines. A prototype model of the power system is created in the laboratory environment in order to test the FPGA-based online intelligent recognition system, determine accuracy rates and evaluate its performance. Power quality event types are created in wide parameter ranges on this model. Obtained results from both recognition systems indicated the hardware and software designs of our embedded systems are quite effective, fast, and have high success performance.

Arrhythmia is a condition characterized by perturbation of the regular rhythm of the heart. The development of computerized self-diagnostic systems for the detection of these arrhythmias is very popular, thanks to the machine learning models included in these systems, which eliminate the need for visual inspection of long electrocardiogram (ECG) recordings. In order to design a reliable, generalizable and highly accurate model, large number of subjects and arrhythmia classes are included in the training and testing phases of the model. In this study, an ECG dataset containing more than 10,000 subject records was used to train and diagnose arrhythmia. A deep neural network (DNN) model was used on the data set during the extraction of the features of the ECG inputs. Feature maps obtained from hierarchically placed layers in DNN were fed to various shallow classifiers. Principal component analysis (PCA) technique was used to reduce the high dimensions of feature maps. In addition to the morphological features obtained with DNN, various ECG features obtained from lead-II for rhythmic information are fused to increase the performance. Using the ECG features, an accuracy of 90.30% has been achieved. Using only deep features, this accuracy was increased to 97.26%. However, the accuracy was increased to 98.00% by fusing both deep and ECG-based features. Another important research subject of the study is the examination of the features obtained from DNN network both on a layer basis and at each training step. The findings show that the more abstract features obtained from the last layers of the DNN network provide high performance in shallow classifiers, and weight updates of DNN network also increases the performance of these classifiers. Hence, the study presents important findings on the fusion of deep features and shallow classifiers to improve the performance of the proposed system.

Human life necessitates high-quality sleep. However, humans suffer from a lower quality of life because of sleep disorders. The identification of sleep stages is necessary to predict the quality of sleep. Manual sleep-stage scoring is frequently conducted through sleep experts' visually evaluations of a patient's neurophysiological data, gathered in sleep laboratories. Manually scoring sleep is a tough, time-intensive, tiresome, and highly subjective activity. Hence, the need of creating automatic sleep-stage classification has risen due to the limitations imposed by manual sleep-stage scoring methods. In this study, a novel machine learning model is developed using dual-channel unipolar electroencephalogram (EEG), chin electromyogram (EMG), and dual-channel electrooculgram (EOG) signals. Using an optimum orthogonal filter bank, sub-bands are obtained by decomposing 30 s epochs of signals. Tsallis entropies are then calculated from the coefficients of these sub-bands. Then, these features are fed an ensemble bagged tree (EBT) classifier for automated sleep classification. We developed our automated sleep classification model using the Sleep Heart Health Study (SHHS) database, which contains two parts, SHHS-1 and SHHS-2, containing more than 8455 subjects with more than 75,000 h of recordings. The proposed model separated three classes if sleep: rapid eye movement (REM), non-REM, and wake, with a classification accuracy of 90.70% and 91.80% using the SHHS-1 and SHHS-2 datasets, respectively. For the five-class problem, the model produces a classification accuracy of 84.3% and 86.3%, corresponding to the SHHS-1 and SHHS-2 databases, respectively, to classify wake, N1, N2, N3, and REM sleep stages. The model acquired Cohen's kappa (κ) coefficients as 0.838 with SHHS-1 and 0.86 with SHHS-2 for the three-class classification problem. Similarly, the model achieved Cohen's κ of 0.7746 for SHHS-1 and 0.8007 for SHHS-2 in five-class classification tasks. The model proposed in this study has achieved better performance than the best existing methods. Moreover, the model that has been proposed has been developed to classify sleep stages for both good sleepers as well as patients suffering from sleep disorders. Thus, the proposed wavelet Tsallis entropy-based model is robust and accurate and may help clinicians to comprehend and interpret sleep stages efficiently.

Laser-induced breakdown spectroscopy (LIBS) is an important analysis technique with applications in many industrial branches and fields of scientific research. Nowadays, the advantages of LIBS are impaired by the main drawback in the interpretation of obtained spectra and identification of observed spectral lines. This procedure is highly time-consuming since it is essentially based on the comparison of lines present in the spectrum with the literature database. This paper proposes the use of various computational intelligence methods to develop a reliable and fast classification of quasi-destructively acquired LIBS spectra into a set of predefined classes. We focus on a specific problem of classification of paper-ink samples into 30 separate, predefined classes. For each of 30 classes (10 pens of each of 5 ink types combined with 10 sheets of 5 paper types plus empty pages), 100 LIBS spectra are collected. Four variants of preprocessing, seven classifiers (decision trees, random forest, k-nearest neighbor, support vector machine, probabilistic neural network, multi-layer perceptron, and generalized regression neural network), 5-fold stratified cross-validation, and a test on an independent set (for methods evaluation) scenarios are employed. Our developed system yielded an accuracy of 99.08%, obtained using the random forest classifier. Our results clearly demonstrates that machine learning methods can be used to identify the paper-ink samples based on LIBS reliably at a faster rate.

In this paper, a new generation Internet-based Power Quality Monitoring System (IPQMS) that transmits real-time power quality (PQ) data over the internet has been developed. The monitoring system has the ability to measure PQ parameters in accordance with the related standards. This monitoring system includes many hardware and software designs and presents an efficient structure. PQ parameters are determined by signal processing algorithms that are applied to the current and voltage signals continuously obtained from the electrical power network. These signal processing algorithms are performed as embedded functions in the FPGA device. The PQ data obtained from the measurement points are transmitted to a server by UDP/IP communication protocol that is implemented in the FPGA device. The monitoring, reporting and permanently storing tasks are accomplished with the real-time automation software and web applications that running on the server computer. With its innovative software and hardware designs, the proposed monitoring approach presents a very useful monitoring structure that can be used in the PQ field.

Blood cell analysis is a crucial diagnostic process in medical practice. In particular, detecting white blood cells (WBCs) is essential for diagnosing of many diseases. The manual screening of blood films is a time-consuming and subjective process, which can lead to inconsistencies and errors. Therefore, automated detection of blood cells can improve the accuracy and efficiency of the screening process. In this study, an explainable Vision Transformer (ViT) model was proposed for the automatic detection of WBCs from blood films. The proposed model utilizes the self-attention mechanism to extract relevant features from the input images and leverages transfer learning by incorporating pre-trained model weights to improve its performance. The proposed model achieved a classification accuracy of 99.40% for five distinct types of WBCs and exhibited potential in reducing the time required for manual screening of blood films by pathologists. Upon examination of the misclassified test samples, it was observed that incorrect predictions were correlated with the presence or absence of granules in the cell samples. To validate this observation, the dataset was divided into two classes, namely Granulocytes and Agranulocytes, and a secondary training process was conducted. The resulting ViT model trained for binary classification achieved an accuracy of 99.70%, recall of 99.54%, precision of 99.32%, and F-1 score of 99.43% during the test phase. To ensure the reliability of the ViT model&amp;#039;s multi-class classification of WBCs, the pixel areas that the model focuses on in its predictions are visualized through the Score-CAM algorithm.

Electrocardiogram (ECG) signals consist of data containing measurements of electrical activity in the heartbeats. These signals include relevant information used to detect abnormalities such as arrhythmia. In this study, a recognition system is proposed for detection and classification of heartbeats in ECG signals. Heartbeats in the ECG data were detected by using the wavelet transform (WT) method and these beats are segmented with determined periods. For obtaining distinctive features from the beats, multi-resolution WT is applied to these segmented signals, and wavelet coefficients are obtained from different frequency levels. Feature vectors are generated on these coefficients by using various statistical methods. The proposed recognition system is trained on feature vectors by using the Online Sequential Extreme Learning Machine (OSELM) classifier during the learning phase to automatically recognize the signals. Five different beat types were obtained from the MIT-BIH arrhythmia dataset. The multi-class dataset that includes five classes and the binary-class dataset that includes two classes were created among these beat types. Performance tests of the proposed wavelet-based-OSELM (W-OSELM) method were realized with these two datasets. The proposed recognition system provided 97.29% correct beat detection rate from raw ECG signals. The classification accuracy is 99.44% for the binary-class dataset and 98.51% for the multi-class dataset. Furthermore, the proposed classifier has shown very fast recognition performance on ECG signals.

This paper presents a recently established compressed sensing (CS) and sparse autoencoder (SAE) based on deep learning (DL) method for classification of single and multiple power quality disturbances (PQDs). The CS technique is paying considerable attention in recent years due to below sampling rate comparatively Nyquist sampling. Initially, the CS technique is applied to extract the features of PQD waveforms. The extracted features are applied as inputs to the sparse autoencoder based on DL for classification of nine single and 22 combined classes of PQDs. The DL helps to remove a redundant feature and improves classification performance. Finally, backpropagation is applied to fine-tune the entire network. The effectiveness of the proposed algorithm has been tested with more than 6580 numbers of real and synthetic single and multiple PQD data, and the results are recorded. High correct classification rate is obtained with noise and without noise level. Noise level was considered from 20 to 50 dB. The performance of the proposed technique has been assessed by comparing the results against recently reported methods. Results show that the proposed CS- and SAE-based DL algorithms can be efficiently used for single and multiple PQDs classifications.

In this paper, a new method based on deep learning has been proposed in order to recognize noise-digital modulation signals at varying noise levels automatically. The 8-bit data from six different modulations have been obtained by adding noise levels from 5 to 25 dB. The used digital modulation types are Amplitude Shift Keying, Frequency Shift Keying, Phase Shift Keying, Quadrature Amplitude Shift Keying, Quadrature Frequency Shift Keying, and Quadrature Phase Shift Keying. To recognize the noise-digital modulation signals automatically, a new deep long short-term memory networks (LSTMs) model has been proposed and then applied to these signals successfully. A significant advantage of the proposed system is that deep learning method has been trained and tested with raw digital modulation signals without applying any feature extraction from the signals. In this study, the noise modulation signals of 5–25 dB have been classified and compared with each other. The innovative aspect of the study is to classify the modulation with the LSTM method without dealing with the extraction of signal characteristics. Without noise, added digital modulation signals had been classified as the success rate of 97.22%, while with all noise-added signals have been classified as the success rate of 94.72% with deep LSTM model. The experimental results show that the proposed deep LSTM model has been achieved remarkable results in recognition of noised six different modulation signals with a fully end-to-end structure.

Prostate cancer (PCa) is the most common type of cancer among men. Digital rectal examination and prostate-specific antigen (PSA) tests are used to diagnose the PCa accurately. Since PSA is organ-specific and not disease-specific, multiparametric magnetic resonance imaging (mpMRI) is used to reduce unnecessary biopsies. Prostate imaging reporting and data system (PI-RADS) is widely used for mpMRI scoring to detect PCa. There is low-level agreement among interpreters and also subjectivity associated with PI-RADS scoring. Hence, in this study, a hybrid model has been proposed to accurately interpret mpMRI examination and predict PI-RADS scores. In the proposed systems, feature maps of mpMR images were extracted using the MobilenetV2, Efficientnetb0, and Darknet53 architectures. Then, the feature maps obtained using these three architectures were combined. The merged feature maps are subjected to neighborhood components analysis (NCA) to eliminate redundant features. The proposed system provided 96.09% accuracy.

Artificial intelligence (AI) has helped to obtain accurate, fast, robust results without any human errors.Hence, it has been used in various applications in our daily lives. The Turing test has been afundamental problem that AI systems aim to overcome. Recently developed various natural language problem (NLP) models have shown significant performances. AI language models, used intranslation, digital assistant, and sentiment analysis, have improved the quality of our lives. It canperform scans on thousands of documents in seconds and report them by establishing appropriatesentence structures. Generative pre-trained transformer (GPT)-3 is a popular model developedrecently has been used for many applications. Users of this model have obtained surprising results onvarious applications and shared them on various social media platforms. This study aims to evaluatethe performance of the GPT-3 model in writing an academic article. Hence, we chose the subject ofthe article as tools based on artificial intelligence in academic article writing. The organized querieson GPT-3 created the flow of this article. In this article, we have made an effort to highlight theadvantages and limitations of using GPT-3 for research paper writing. Authors feel that it can be usedas an adjunct tool while writing research papers.

Abstract The automated segmentation of dental restorations is a critical step in diagnosing dental problems and suggesting the best treatment. Some restorations may be missed during a dental examination, depending on the number of patients, the dentist's experience, and fatigue. Automatic detection of dental restorations based on deep learning has the potential to provide a quick radiological assessment based on the patient's treatment history and pre‐diagnosis. This study presents a deep learning‐based method for automatic detection and classification of amalgam and composite fillings on panoramic images. A total of 250 anonymized panoramic images with amalgam and composite fillings with a resolution of 2048 × 1024 px were used. In this study, U‐Net models with various backbones were employed. The ResNext50 model has achieved the highest pixel accuracy and intersection over union (IoU) performance based on the evaluation of various ResNet and ResNext backbones. The mean IoU value obtained by the model on the test images is 0.767 while the Pixel Accuracy of 99.81% was achieved. Our proposed method demonstrated superior performance compared to similarly conducted studies in the literature. The proposed method can potentially be employed in clinical settings to detect dental restorations automatically. The classification and detection of dental restorations with this model can aid dentistry education at higher institutions as an education tool and make the reporting easier for the dentist.

Thanks to the technologies that have become a part of daily life, huge data piles are formed in almost every field. Research on detecting hidden patterns in big data and discovering useful information has gained importance. The amount of data accumulated in the field of education has enabled data mining techniques to come to the fore in this field as an alternative to traditional statistical methods. In traditional statistical methods, hidden relationships between some variables can be ignored. This can cause some information to be lost or not to use essential data in essential areas such as education. However, educational data mining (EDM) can unlock valuable data and predict important relationships to improve and improve the quality of education. For this reason, this study aimed to perform a sample EDM application to draw attention to its EDM predictive power. The data set consisted of the opinions collected from university students. This data set variables were formed by distance education students' academic self-efficacy, sense of community, academic achievement averages, and some demographic variables. The descriptive model revealed latent patterns between variables in the study, and a predictive model was used to estimate variables. For this, the association rule method and classification algorithm were also used. At the end of the study, it was concluded that EDM could effectively find relationships between variables and predict variables.

Background and objective: Deep learning structures have recently achieved remarkable success in the field of machine learning. Convolutional neural networks (CNN) in image processing and long-short term memory (LSTM) in the time-series analysis are commonly used deep learning algorithms. Healthcare applications of deep learning algorithms provide important contributions for computer-aided diagnosis research. In this study, convolutional long-short term memory (CLSTM) network was used for automatic classification of EEG signals and automatic seizure detection. Methods: A new nine-layer deep network model consisting of convolutional and LSTM layers was designed. The signals processed in the convolutional layers were given as an input to the LSTM network whose outputs were processed in densely connected neural network layers. The EEG data is appropriate for a model having 1-D convolution layers. A bidirectional model was employed in the LSTM layer. Results: Bonn University EEG database with five different datasets was used for experimental studies. In this database, each dataset contains 23.6[Formula: see text]s duration 100 single channel EEG segments which consist of 4097 dimensional samples (173.61[Formula: see text]Hz). Eight two-class and three three-class clinical scenarios were examined. When the experimental results were evaluated, it was seen that the proposed model had high accuracy on both binary and ternary classification tasks. Conclusions: The proposed end-to-end learning structure showed a good performance without using any hand-crafted feature extraction or shallow classifiers to detect the seizures. The model does not require filtering, and also automatically learns to filter the input as well. As a result, the proposed model can process long duration EEG signals without applying segmentation, and can detect epileptic seizures automatically by using the correlation of ictal and interictal signals of raw data.

White blood cells (WBCs) are crucial components of the immune system that play a vital role in defending the body against infections and diseases. The identification of WBCs subtypes is useful in the detection of various diseases, such as infections, leukemia, and other hematological malignancies. The manual screening of blood films is time-consuming and subjective, leading to inconsistencies and errors. Convolutional neural networks (CNN)-based models can automate such classification processes, but are incapable of capturing long-range dependencies and global context. This paper proposes an explainable Vision Transformer (ViT) model for automatic WBCs detection from blood films. The proposed model uses a self-attention mechanism to extract features from input images. Our proposed model was trained and validated on a public dataset of 16,633 samples containing five different types of WBCs. As a result of experiments on the classification of five different types of WBCs, our model achieved an accuracy of 99.40%. Moreover, the model's examination of misclassified test samples revealed a correlation between incorrect predictions and the presence or absence of granules in the cell samples. To validate this observation, we divided the dataset into two classes, Granulocytes and Agranulocytes, and conducted a secondary training process. The resulting ViT model, trained for binary classification, achieved impressive performance metrics during the test phase, including an accuracy of 99.70%, recall of 99.54%, precision of 99.32%, and F-1 score of 99.43%. To ensure the reliability of the ViT model's, we employed the Score-CAM algorithm to visualize the pixel areas on which the model focuses during its predictions. Our proposed method is suitable for clinical use due to its explainable structure as well as its superior performance compared to similar studies in the literature. The classification and localization of WBCs with this model can facilitate the detection and reporting process for the pathologist.

Software development processes are continuously evolving and rapidly transforming alongside the rapid changes in technology. Recently, innovations in the field of Artificial Intelligence (AI) have led to significant changes in software development practices. AI tools can greatly enhance traditional software development processes by offering developers the ability to create projects more intelligently, swiftly, and effectively. These tools can be employed in various tasks, such as code generation, test automation, error analysis, and performance improvements. Particularly, ChatGPT, an AI-based language model that has had a profound impact on almost every domain, can assist software developers in writing code faster and in a more natural language manner. In this research article, essential information about the usage of ChatGPT in the software development process is presented. To evaluate some capabilities of ChatGPT in the software development context, applications were performed on a software project. For this purpose, a software development process was constructed based on the responses provided by ChatGPT. Various questions related to software development processes were formulated, and the responses generated by GPT were evaluated. The obtained results indicated that ChatGPT exhibited excellent performance in the software development process. Based on these findings, it was observed that AI-based models like ChatGPT could be effectively utilized as assisting tools in software development processes, accelerating traditional workflows. Furthermore, AI-based tools can automate testing processes, enhancing software quality while saving time and effort.

Malware detection and classification have become critical tasks in ensuring the security and integrity of computer systems and networks. Traditional methods of malware analysis often rely on signature-based approaches, which struggle to cope with the ever-evolving landscape of malware variants. In recent years, deep learning techniques have shown promising results in automating the process of malware classification. This paper presents a novel approach to malware image classification using the Vision Transformer (ViT) architecture. In this work, we adapt the ViT model to the domain of malware analysis by representing malware images as input tokens to the ViT architecture. To evaluate the effectiveness of the proposed approach, we used a comprehensive dataset comprising 14,226 malware samples across 26 families. We compare the performance of our ViT-based classifier with traditional machine learning methods and other deep learning architectures. Our experimental results showcase the potential of the ViT in handling malware images, achieving a classification accuracy of 98.80%. The presented approach establishes a strong foundation for further research in utilizing state-of-the-art deep learning architectures for enhanced malware analysis and detection techniques.

To develop effective advanced driving assistance systems, it is important to accurately recognize current driving environments and make critical decisions about driving processes.Preventing accidents through the interaction between the driving assistance systems and the environment and ensuring optimum driving dynamics are the main topics in this field.Vehicles need to recognize the road type and quality at a high accuracy to ensure the most suitable driving for the road type.It is also important to use both uncomplicated and cost-effective systems when performing this detection.In this study, a deep learning-based approach that can be used in vehicle driver assistance systems is proposed to automatically recognize road type and quality.Using this approach, it is possible to determine the road type and the quality of the road using only driving images as the input data.A new convolutional neural network model is designed for classification of the driving images.Driving images obtained from Google Street View are used to evaluate the recognition system for an actual driving environment.The proposed approach shows that the road types were determined with accuracy of 91.41 %, and the pothole road-smooth road distinction was successful at 91.07 %.It can be said that the proposed method is an effective structure that can be used for advanced driving support systems, V2I communications systems, and similar intelligent transportation systems.

COVID-19 caused by the transmission of SARS-CoV-2 virus taking a huge toll on global health and caused life-threatening medical complications and elevated mortality rates, especially among older adults and people with existing morbidity. Current evidence suggests that the virus spreads primarily through respiratory droplets emitted by infected persons when breathing, coughing, sneezing, or speaking. These droplets can reach another person through their mouth, nose, or eyes, resulting in infection. The “gold standard” for clinical diagnosis of SARS-CoV-2 is the laboratory-based nucleic acid amplification test, which includes the reverse transcription-polymerase chain reaction (RT-PCR) test on nasopharyngeal swab samples. The main concerns with this type of test are the relatively high cost, long processing time, and considerable false-positive or false-negative results. Alternative approaches have been suggested to detect the SARS-CoV-2 virus so that those infected and the people they have been in contact with can be quickly isolated to break the transmission chains and hopefully, control the pandemic. These alternative approaches include electrochemical biosensing and deep learning. In this review, we discuss the current state-of-the-art technology used in both fields for public health surveillance of SARS-CoV-2 and present a comparison of both methods in terms of cost, sampling, timing, accuracy, instrument complexity, global accessibility, feasibility, and adaptability to mutations. Finally, we discuss the issues and potential future research approaches for detecting the SARS-CoV-2 virus utilizing electrochemical biosensing and deep learning.

Disturbances occurring on voltage signals following switching events or faults on power systems are defined as power quality problem. Power quality problems result in significant financial losses in the power systems in addition to faulty operation or breakdown of sensitive loads connected to the power system. Therefore, power quality problems should be swiftly detected and eliminated. This paper proposes a real-time embedded detection system for the purpose of identification of power quality disturbances. The detection system developed has been created using the wavelet transform in the Field Programmable Gate Array (FPGA) media. The general structure of FPGA-based detection system comprises two fundamental parts: hardware and software. Hardware structure comprises signal input card, FPGA and computer components. The software stage includes FPGA and graphical interface software. An experimental setup of the power system has been created in laboratory environment in order to test the FPGA-based detection system, determine accuracy rates and assess its success. Various power quality disturbances have been created over this model in wide parameter intervals. When the results obtained for the real-time power quality disturbance detection system developed under this study are examined, it has been seen that hardware and software designs are quite effective, fast and of high achievement performance.

Elektrik motorları, çeşitli işlemleri otomatikleştirme ve kolaylaştırma yeteneklerinden dolayı endüstride önemli bir yere sahiptir. Elektrik motorlarında meydana gelen arızalar, cihazın veya sistemin çalışmasını etkileyebilmekte ve büyük maddi kayıplara neden olabilmektedir. Bu nedenle arızaların erken aşamada tespit edilmesi kritik bir öneme sahiptir. Arızaların tespitinde bilgisayar destekli yazılımlar kullanılması maliyetten ve zamandan tasarruf etme potansiyeli nedeniyle ön plana çıkmaktadır. Bu çalışmada, motor yatağı arıza türlerini tespit etmek için derin öğrenme tabanlı bir model önerilmiştir. Tek boyutlu konvolüsyonel sinir ağı (1D-CNN) mimarisi kullanan bu model ile sadece titreşim verileri kullanılarak arıza tipi tespiti sağlanmaktadır. Önerilen mimari, titreşim sinyallerini motor arıza teşhisinde hızlı ve güvenilir olarak kullanan etkin bir modeldir. Çalışma kapsamında farklı hız senaryoları kullanılarak eğitim ve test aşamalarının detaylı performans değerlendirmeleri sağlanmıştır. Genelleme kabiliyeti yüksek olan bu model ile, farklı senaryolarda yüksek doğruluk oranları ile arıza tespiti yapılmıştır.

In this paper, the Convolutional Neural Network (CNN) architecture, which is one of the deep learning architectures, is used to classify the basic circuit components drawn by hand. During the training and testing stages of the model, a new dataset containing images of 863 circuit components manually drawn by different people is created. The data set contains images of four different classes of circuit components such as resistor, inductor, capacitor and voltage source. All images have been fixed to the same size and converted to grayscale to increase recognition performance and reduce process complexity. In the study, training for four classes is performed with CNN architecture. Based on the CNN architecture, four new CNN models are employed with different the number of layers. The training and validation results of these models are compared separately, the model with the highest training and validation performance is observed with four layer CNN model (CNN-4). This model obtained 84.41% accuracy rate at classification task.

In this study, a deep learning method was used to detect sleep states of the drivers in the driving environment. A convolutional neural network (CNN) model has been proposed to determine whether the eyes of certain constant face images of drivers are closed. The proposed model has a wide potential application area such as human-computer interface design, facial expression recognition, driver fatigue-sleepiness determination. This method, which was developed on driver sleepiness data, has been applied on 4,846 real eye images in the Closed Eyes In The Wild (CEW) database. Commonly used CNN models are used on the same data to compare performances of the prepared model. According to the classification results obtained, 96.5% and 92.99% of the designed model achieved success and it is seen that this structure can be used in this problem area.

Images obtained from the real world environments usually have various distortions in image quality. For example, when an object in motion is filmed, or when an environment is being filmed on the move, motion tracking effects occur on the image. Increasing the recognition performance of expert systems, which perform image recognition on data obtained under such conditions, is an important research area. In this study, we propose a Convolutional Neural Network (CNN) based Deep System Model (CNN-DSM) for accurate classification of images under challenging conditions. In the proposed model, a new layer is designed in addition to the classical CNN layers. This layer works as an enhancement layer. For the performance evaluations, various real world surface images were selected from the Curet database. Finally, results are presented and discussed.

In this study, a feature vector optimization based method has been proposed for classification of the heartbeat types. Electrocardiogram (ECG) signals of five different heartbeat type were used for this aim. Firstly, wavelet transform (WT) method were applied on these ECG signals to generate all feature vectors. Optimizing these feature vectors is provided by performing particle swarm optimization (PSO), genetic search, best first, greedy stepwise and multi objective evoluationary algorithms on these vectors. These optimized feature vectors are later applied to the classifier inputs for performance evaluation. A comprehensive assessment was presented for the determination of optimized feature vectors for ECG signals and best-performing classifier for these optimized feature vectors was determined.

In this paper, a new 2D-DOST (Two-Dimensional Discrete Orthonormal Stockwell Transform) and LS-SVM (Least Squares Support Vector Machines) based classifier system is proposed for classification of texture images. The proposed system contains two main stages. These stages are feature extraction and classification. In the feature extraction stage, the distinguishing feature vectors which represent descriptive features of texture images are obtained by using a 2D-DOST based feature extraction method. In the classification stage, the texture images are classified by the LS-SVM since this classifier has high success rate and accuracy. The training of LS-SVM is performed on the distinguishing feature vector of each texture component. Texture samples are recognized by the test data applied to the input of trained LS-SVM classifier. Performance evaluations of the proposed method are carried on different datasets obtained from sub-images. These datasets include both the normal texture images and noise added images. Sub-images into datasets are derived from Brodatz and Kylberg texture images database. Gaussian and Salt & Pepper noise with different levels are used for creating noisy datasets. According to the study results, the proposed 2D-DOST and LS-SVM based classifier has a capability of classifying texture images with high success rate and noise robustness.

In this paper, an effective classification approach to classify harmonic data has been proposed. In the proposed classifier approach, harmonic data obtained through a 3-phase system have been classified by using $k$-means and least square support vector machine (LS-SVM) models. In order to obtain class details regarding harmonic data, a $k$-means clustering algorithm has been applied to these data first. The training of the LS-SVM model has been realized with the class details obtained through the $k$-means algorithm. To increase the efficiency of the LS-SVM model, the regularization and kernel parameters of this model have been determined with a grid search method and the training phase has been realized. Backpropagation neural network and J48 decision tree classifiers have been applied to the same data and results have been obtained for the purpose of comparing the performance of the LS-SVM model. The real data obtained from the output of distribution system have been used to assess the performance of the proposed classifier system. The obtained results and comparisons suggest that the proposed classifier system approach is quite efficient at classifying harmonic data.

Hypertension (HTN) is a major risk factor for cardiovascular diseases. At least 45% of deaths due to heart disease and 51% of deaths due to stroke are the result of hypertension. According to research on the prevalence and absolute burden of HTN in India, HTN positively correlated with age and was present in 20.6% of men and 20.9% of women. It was estimated that this trend will increase to 22.9% and 23.6% for men and women, respectively, by 2025. Controlling blood pressure is therefore important to lower both morbidity and mortality. Computer-aided diagnosis (CAD) is a noninvasive technique which can determine subtle myocardial structural changes at an early stage. In this work, we show how a multi-resolution analysis-based CAD system can be utilized for the detection of early HTN-induced left ventricular heart muscle changes with the help of ultrasound imaging. Firstly, features were extracted from the ultrasound imagery, and then the feature dimensions were reduced using a locality sensitive discriminant analysis (LSDA). The decision tree classifier with contourlet and shearlet transform features was later employed for improved performance and maximized accuracy using only two features. The developed model is applicable for the evaluation of cardiac structural alteration in HTN and can be used as a standalone tool in hospitals and polyclinics.

In this study, an FPGA based online monitoring system was developed for detection of power quality disturbances. The developed system has the capability to instantly detect commonly seen PQ disturbances such as voltage sag, voltage swell and interruption occurring on real-time three-phase voltage signals obtained from the electric power system. The data of the detected disturbance events are transferred to the visual software prepared in the computer environment by means of the UDP/IP communication module embedded in the FPGA. This PQ disturbance detection system was tested upon installation at three different measuring points across the campus of Tunceli University. It was seen from the test results examined that the detection system successfully identified the PQ disturbance events.

Abstract This study aimed to perform a sample educational data mining (EDM) application to draw attention to its EDM predictive power. The data set consisted of the opinions collected from university students. These data set variables were formed by distance education students' academic self-efficacy, sense of community, academic achievement averages, and some demographic variables. The descriptive model revealed latent patterns between variables in the study, and a predictive model was used to estimate variables. For this, the association rule method and classification algorithm were also used. At the end of the study, it was concluded that EDM could effectively find relationships between variables and predict variables.

Breast cancer is a type of cancer caused by the uncontrolled growth and proliferation of cells in the breast tissue. Differentiating between benign and malignant tumors is critical in the detection and treatment of breast cancer. Traditional methods of cancer detection by manual analysis of radiological images are time-consuming and error-prone due to human factors. Modern approaches based on image classifier deep learning models provide significant results in disease detection, but are not suitable for clinical use due to their black-box structure. This paper presents a semantic segmentation method for breast cancer detection from ultrasound images. First, an ultrasound image of any resolution is divided into 256×256 pixel patches by passing it through an image cropping function. These patches are sequentially numbered and given as input to the model. Features are extracted from the 256×256 pixel patches with pre-trained ResNext models placed in the encoder network of the U-Net model. These features are processed in the default decoder network of the U-Net model and estimated at the output with three different pixel values: benign tumor areas (1), malignant tumor areas (2) and background areas (0). The prediction masks obtained at the output of the decoder network are combined sequentially to obtain the final prediction mask. The proposed method is validated on a publicly available dataset of 780 ultrasound images of female patients. The ResNext-based U-Net model achieved 73.17% intersection over union (IoU) and 83.42% dice coefficient (DC) on the test images. ResNext-based U-Net models perform better than the default U-Net model. Experts could use the proposed pixel-based segmentation method for breast cancer diagnosis and monitoring.

Electrical machines, which provide many conveniences in our daily life, may experience malfunctions that may adversely affect their performance and the general functioning of the industrial processes in which they are used. These failures often require maintenance or repair work, which can be expensive and time consuming. Therefore, minimizing the risk of malfunctions and failures and ensuring that these machines operate reliably and efficiently play a critical role for the industry. In this study, a one-dimensional convolutional neural network (1D-CNN) based fault diagnosis model is proposed for electric motor fault detection. Motor vibration data was chosen as the input data of the 1D-CNN model. Motor vibration data was obtained from a mobile application developed by using the three-axis accelerometer of the mobile phone. Three-axis data (X-axis, Y-axis and Z-axis) were fed to the model, both separately and together, to perform motor fault detection. The results showed that even a single axis data provides error-free diagnostics. With this fault detection method, which does not require any connection on or inside the motor, the fault condition in an electric motor has been detected with high accuracy.

A brain stroke occurs when blood flow to a part of the brain is disrupted, potentially caused by a blocked or ruptured blood vessel. Deprived of oxygen and nutrients, brain cells can start dying within minutes, leading to irreversible damage. Early diagnosis and treatment are crucial to minimize brain damage and improve recovery chances. Clinical assessments and imaging techniques like Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scans are commonly used for rapid detection, but manual analysis has limitations, including delays and subjectivity. AI-based models offer a faster and more consistent approach for stroke diagnosis, enhancing accuracy. In this study, an explainable Vision Transformer (ViT) model is proposed for stroke classification and localization from brain CT images. The model is validated on a dataset of 6,651 samples. To address an unbalanced dataset, two training scenarios were employed. Scenario-1 directly used the unbalanced dataset, while Scenario-2 equalized sample numbers through data augmentation. In the test phase, Scenario-1 achieved 97.25% accuracy, 98.46% precision, 96.00% recall, 98.50% specificity, and a 97.22% F-1 score. In contrast, Scenario-2 achieved even higher performance with 98.75% accuracy, 99.49% precision, 98.00% recall, 99.50% specificity, and a 98.74% F-1 score. Analyzing the softmax ratios in the model predictions revealed that Scenario-2, with synthetic images in the training set, produced more reliable results. The study also used the Grad-CAM algorithm to visualize the areas of focus in the models’ predictions, showcasing their superior localization capabilities. This proposed model is well-suited for clinical use due to its high accuracy rates and robust localization abilities, potentially improving stroke diagnosis and treatment outcomes.

Clinical screening of myocardial infarction is important for preventative treatment and risk stratification in cardiology practice, however current detection by electrocardiogram Q-wave analysis provides only modest accuracy for assessing prior cardiac events. We set out to evaluate the ability of a deep neural network trained on the electrocardiogram to identify patients with clinical history of myocardial infarction. We assessed 608 patients at two academic centers with adjudicated history of myocardial infarction. Surface electrocardiograms were used to train a neural network-based model that classifies patients with and without a history of infarction. Endpoints were assessed by clinical record review and accuracy of the model was compared against the manual assessment of pathologic Q waves. The neural network outperformed the accuracy of pathologic Q waves (62%). In training, the model accuracy converged to >98%. Validation was performed by cross-validation (k=5) with validation accuracy 71 ± 5%. Receiver-operator characteristics analysis resulted in a c-statistic of 0.730. Deep learning of a 12-lead ECG can identify features of prior myocardial injury more accurately than clinical Q-wave analysis and may serve as a valuable clinical screening tool.

In robotics, manipulators are recently becoming one of the prominent fields of interest for different types of applications. One of the usual functionalities performed by manipulators is grasping. Grasping means simply holding an object. In order to perform a grasping task, each manipulator needs a gripper mounted at the end effector of them. In this paper, a method based on deep reinforcement learning is presented to deal with the issue of robotic grasping employing only vision feedback. The combination of deep learning with dueling architecture, a variant of Q-learning, brings the complexity caused by the use of handcrafted features to a humbler state. Our method employs the Dueling Deep Q-learning Network(DDQN) to learn the grasping policy. Our proposed system employs a visual structure that uses a Kinect camera setup that spots the scene that possesses the object of interest. We realized our experiments by utilizing Webots simulator environment. The results show that our proposed dueling architecture enables our Reinforcement Learning(RL) agent to perform well enough to fulfill the grasping task.

In this study, a real-time monitoring system for the purpose of monitoring 3-phase voltage harmonics instantaneously has been prepared by using Field Programmable Gate Arrays (FPGA). In the monitoring system that has been achieved, 3-phase voltage signals are transferred with a devised signal input card to an FPGA device. The harmonic values of the voltage signals belonging to each phase are obtained instantaneously with the 128 floating-point format FFT algorithm embedded in an FPGA. The obtained harmonic values are transferred onto a computer medium with the communication protocol RS232 that is created in the FPGA device. A software has been created for the purpose of visual monitoring of both the graphical interface program that has been done on computer and the harmonic data belonging to these signals.

In this study, a new FPGA based monitoring system has been developed for monitoring the power quality. Designed system performs the calculations of power quality parameters on real-time voltages and currents datas that are obtained from the network. FPGA devices that have recently become more popular in digital signal processing field is used in the structure of monitoring system. Designed power quality monitoring system has been established on electrical distribution panels of Tunceli University Vocational Schools. When the obtained test results are evaluated, it has been observed that the proposed power quality monitoring system is successful.

Extreme learning machine (ELM) and deep learning methods are well-known with their efficiency, accuracy, and speed. In this study, we focus on the application of ELM to a deep learning structure for person recognition with facial expressions. For this purpose, a new convolutional neural network (CNN) model containing Kernel ELM classifiers was constructed. In this model, ELM was not used only as a fully connected layer replacement and energy function was employed to generate feature maps for the ELM. There are two advantages of the proposed model. First, it is fast and successful in face recognition studies. Second, it can drastically improve the performance of a partially-trained CNN model. Consequently, the proposed model is very suitable for CNN models, where the learning process requires a lot of time and computational power. The model is tested with the Grimace data set and experimental results are presented in details.

Deep learning algorithms have recently become the most widely used machine learning approaches. Remarkable approaches have emerged in the field of machine learning studies with the use of geometric methods for deep learning. The manifold theory, which is simply regarded as the general state of an Euclidean space, has become applicable in the field of deep learning and this field has been named Geometric Deep Learning. In this study, applications of manifold theory and differential geometry to deep learning is presented.

In this paper, a classification system based on association rule to determine the types of power quality event is presented. Firstly, a single feature vector representing three phase event signal is obtained by applying the wavelet transform to event signals in this system. The inputs of generating association rules algorithm are obtained by applying proper transform process to these feature vectors. Later, obtained rules and support values belong to these rules are stored in a database and used for classification process. Then, power quality events obtained from ATP/EMTP software are applied to the proposed classification system. The results showed that proposed system has high classification accuracy.

Interpretation of electrocardiography (ECG) signals is required for diagnosing cardiac arrhythmia. Recently, machine learning techniques have been applied for automated computer-aided diagnosis. Machine learning tasks can be divided into regression and classification. Regression can be used for noise and artifacts removal as well as resolve issues of missing data from low sampling frequency. Classification task concerns the prediction of output diagnostic classes according to expert-labeled input classes. In this work, we propose a deep neural network model capable of solving regression and classification tasks. Moreover, we combined the two approaches, using unlabeled and labeled data, to train the model. We tested the model on the MIT-BIH Arrhythmia database. Our method showed high effectiveness in detecting cardiac arrhythmia based on modified Lead II ECG records, as well as achieved high quality of ECG signal approximation. For the former, our method attained overall accuracy of 87:33% and balanced accuracy of 80:54%, on par with reference approaches. For the latter, application of self-supervised learning allowed for training without the need for expert labels. The regression model yielded satisfactory performance with fairly accurate prediction of QRS complexes. Transferring knowledge from regression to the classification task, our method attained higher overall accuracy of 87:78%.

Bu calismada, elektrik enerji kalitesi parametrelerini cevrimici izleyen yeni bir enerji kalite izleme cihazi gelistirilmistir. Izleme cihazi, baglanti noktalarindan uc faz akim ve gerilim bilgilerini gercek zamanli elde ederek guc kalitesi parametrelerini hesaplamaktadir. Ayrica sahip oldugu haberlesme yapisi ile elde ettigi olcum bilgilerini uzak baglanti noktalarina anlik olarak aktarmaktadir. Cihaz yapisinda, sinyal isleme alaninda gittikce yayginlasan Alanda Programlanabilir Kapi Dizileri (APKD) donanimi kullanilmistir. Gomulu olarak hazirlanan enerji izleme cihazi, kendi enerji akislarini denetim altinda tutmak isteyen kullanicilarin kolaylikla kullanabilecegi etkin bir platform sunmaktadir.

Bu makale calismasinda, dis ortam goruntulerinde yer alan insan hareketlerinin otomatik siniflandirilmasi icin hibrit bir derin ogrenme yaklasimi onerilmistir. Ilk olarak, dis ortamdan cekilen goruntu icerisindeki kisilerin tespiti saglanmistir. Bu amacla, literaturde yaygin olarak kullanilan onceden egitilmis derin nesne tespit araci olan YOLO kullanilmistir. Dis ortam goruntulerinin elde edilmesinde Google Street View platformu tercih edilmistir. Daha sonra tespit edilen kisiler icin hareket siniflari olusturulmustur. Bu hareket siniflari; saga yurume, sola yurume, ayakta durma ve oturma seklindedir. Boylece dis ortam goruntulerinden tespit edilen kisiler icin kapsamli bir veri seti olusturulmustur. Siniflari belirlenen verilerin otomatik olarak taninmasi islemi icin bir konvolusyonel sinir agi (KSA) modeli tasarlanmistir. Egitimi tamamlanan bu model, YOLO nesne tespit sistemi ile hibrit bir sekilde kullanilarak giris goruntusu icerisindeki kisi hareketlerinin otomatik olarak taninmasini saglamistir. Makale kapsaminda, dort sinifli bir veri seti olusturularak onerilen sistemin performans degerlendirmeleri yapilmistir.

With the advance of nanotechnology, the demand for very dense nanopatterns over large areas continuously increases. Electron beam lithography is one of the main tools to realise this type of nanopattern. In this study we investigate the effect and interrelation of dose and development time for electron beam resist poly(methyl methacrylate) (PMMA) for the realisation of very large area and very dense nanopatterns. For this purpose a pattern consisting of a lattice of holes with radius of 80 nm and lattice constant of 200 nm has been designed, electron beam lithography is realised on a thin resist and a detailed SEM image inspection of the pattern is recorded. Hough transform, a pattern recognition method for the digital images, is applied to SEM images to measure the radii of the holes on the images. Interrelation between dose and development time is investigated. It is found that for a very large area of dense nanopatterns the lithographic process works better for development times shorter than typically used ones.

Repeated neural networks (TSA) have recently been used in the field of machine learning with the memory structure they have. Long-short term memory (UKVH) models, which are an improvement of these networks, are commonly used in the evaluation of sequential data. In this study, the use of UKVH networks in regional earthquake estimation is provided. For this purpose, a five-layer deep UKVH network model was prepared. Bingol in Turkey for the application of this model is based around this central province in the central 50 km radius of a circular area of earthquake prediction has been established. Continuous earthquakes in this circular area are used in the training of the UKVH network. It is provided that the network that has completed the training will predict future earthquakes on the same region. The results were discussed.

Laser-induced breakdown spectroscopy (LIBS) is an important analysis technique with applications in many industrial branches and fields of scientific research. Nowadays, the advantages of LIBS are impaired by the main drawback in the analysis of collected data. This procedure is essentially based on the comparison of lines present in the spectrum with a literature database. This paper proposes the use of various computational intelligence methods to develop a reliable and fast classification of non-destructively acquired LIBS spectra into a set of predefined classes. We focus on a specific problem of classification of paper-ink samples into 30 separate, predefined classes. For each of 30 classes (10 pens of each of 5 ink types combined with 10 sheets of 5 paper types plus empty pages) 100 LIBS spectra are collected. Four variants of preprocessing, seven classifiers (Decision trees, Random forest, k-Nearest Neighbour, Support Vector Machine, Probabilistic Neural Network, Multi-Layer Perceptron, and Generalized Regression Neural Network), 5-fold stratified cross-validation and test on an independent set (for methods evaluation) scenarios are employed. Our developed system yielded an accuracy of 99.08% with average classification time of about 0.12 s is obtained using the random forest classifier. Our results clearly demonstrates that machine learning methods can be used to identify the paper-ink samples based on LIBS reliably at a faster rate.

It is an important work area to determine realtime characteristics of roads where vehicles are in motion in critical areas where artificial intelligence is effectively used, such as driverless vehicles. The purpose of this article work is to present a deeper learning method that will allow a vehicle in motion to detect the direction of flow in the path. Convolutional Neural Networks (KSA) have been used as deep learning models for the determination of the direction of flow (YAY) in the study. The YAY-KSA model developed for flow direction detection is applied on 587 real road images in the CMU VASC image database. To compare the performances of the prepared model, Cifar model which is a common KSA model was applied on the same data. According to the classification results obtained, it was seen that the designed YAY-KSA model correctly determined flow direction at 80.1% level.

Atriyal Fibrilasyon (AFib), yaşlılarda ve hatta herhangi bir kalp hastalığı olmayan gençlerde bile görülebilen yaygın bir kalp ritim bozukluğudur. AFib; inme, kalp yetmezliği ve ani ölümlere neden olabilir. Tüm bu mevcut ve gelecekteki endişeler, dünya çapında AFib'in erken tespitinde önemli önlemlerin alınmasını gerektirir. Elektrokardiyografik (EKG) dalga formları, AFib gibi anormal kalp ritimlerini saptamak için en güvenilir yöntem olarak kabul edilmektedir. Ancak EKG sinyallerinin karmaşıklığı ve doğrusal olmaması nedeniyle bu sinyalleri manuel olarak analiz etmek zordur. Bunun yanı sıra, EKG sinyallerinin yorumlanması kişiye özgü ve uzmanlar arasında farklılık gösterebilmektedir. Bu nedenle otomatik ve güvenilir bir AFib algılama için bilgisayar destekli teşhis (BDT) sistemlerinin kullanımı önemlidir. BDT sistemleri, EKG sinyallerinin değerlendirilmesinin objektif ve doğru olmasını sağlayacak potansiyele sahiptir. Bu çalışmada, derin öğrenme yapısı kullanılarak EKG sinyallerinden otomatik AFib tespiti gerçekleştirilmiştir. Derin öğrenme algoritmalarından evrişimli sinir ağı (ESA) mimarisinin AFib sınıflandırma probleminde kullanımı için çalışma kapsamında derin bir ağ modeli tasarlanmıştır. Kullanılan verisetinde normal sinüs ritimlerinin (SR) yanısıra AFib ve Atriyal Flutter (AFL) aritmileri bulunmaktadır. AFib ve AFL sınıfları birleştirilerek model çıkışında SR ve AFib ayırımının otomatik yapılması sağlanmıştır. Önerilen model, 2222 SR ve 2218 AFib tanısı alan kişilere ait her biri 5000 örneğe sahip EKG sinyali içeren veri seti üzerinde uygulanmıştır. Çalışma kapsamında hazırlanan ESA modeli, test aşamasında sırasıyla %95.09 hassasiyet, %97.27 özgüllük ve %97.26 kesinlik değerlerine ulaşmıştır. Modelin test verileri üzerindeki doğruluk oranı %96.17 olarak elde edilmiştir.