Yazar "Şentürk, Ümit" seçeneğine göre listele
Listeleniyor 1 - 12 / 12
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe Classification of freezing of gait in Parkinson's disease using machine learning algorithms(Institute of Electrical and Electronics Engineers Inc., 2023) Önder, Mithat; Şentürk, Ümit; Polat, Kemal; Paulraj, D.Freezing of gait (FoG) is a prevalent and incapacitating symptom that affects individuals diagnosed with Parkinson's disease (PD) and other movement disorders. Detecting FoG is crucial for accurate diagnosis, fall prevention, and providing objective measurements, all of which are essential for optimizing treatment strategies and improving the quality of life for individuals with FoG. In this study, FoG has been detected using three different classification algorithms: Medium Gaussian Support Vector Machine (SVM), Medium K-Nearest Neighbor (KNN), and Boosted Trees. The process starts with data segmentation, where the dataset is divided into smaller segments. Then, feature extraction is performed on each segment to obtain various statistical measures such as mean, root mean square, maximum, standard deviation, kurtosis, skewness, and peak of root mean square. To ensure a robust and reliable analysis, the dataset is resampled using bootstrapping, a statistical technique that involves drawing random samples from the dataset with replacement. This leads to a more representative sample and reduces the impact of outliers or imbalanced data. The next step is to split the resampled dataset into three different approaches for the classification algorithm: In 5-FCV, the dataset is divided into five equal-sized subsets. Similarly, 10-FCV splits the dataset into ten subsets and follows the same process. Finally, the Medium Gaussian SVM, Medium KNN, and Boosted Trees classification algorithms are applied to the FoG dataset. The classification accuracy achieved is 86.9%, 87.6%, and 92.7% with 10-fold cross-validation, indicating that these algorithms are effective in accurately classifying FoG. © 2023 IEEE.Öğe Cuff-less continuous blood pressure estimation from Electrocardiogram(ECG) and Photoplethysmography (PPG) signals with artificial neural network(Institute of Electrical and Electronics Engineers Inc., 2018) Şentürk, Ümit; Yücedağ, İbrahim; Polat, KemalContinuous blood measurement important information about the health status of the individuals. Conventional methods use a cuff for blood pressure measurement and cannot be measured continuously. In this study, we proposed a system that estimates systolic blood pressure (SP) and diastolic blood pressure (DP) for each heart beat by extracting attributes from ECG and PPG signals. Simultaneous ECG and PPG signals from the PhysioNet Database are pre-processed (denoising, artifact cleaning and baseline wandering) to remove noise and artifacts and segmented into R-R peaks. For each heartbeat, 22-time domain features were extracted from ECG and PPG signals. SP and DP values were estimated by introducing these 22 attributes to the model of Lavenberg-Marquardt artificial neural networks (ANN). Arterial blood pressure (ABP) was also taken from the PhysioNet MIMIC II database along with ECG and PPG signals. ABP signals have been used as targets in the artificial neural network. The system performance has been evaluated by calculating the difference between the estimated ABP values and the actual by the ANN model. The performance value between the predicted SP and actual SP values is -0.14 ± 2.55 (mean ± standard deviation) and the performance value between estimated DP and actual DP values is -0.004 ± 1.6. The obtained results have shown that the proposed model has predicted blood pressure with high accuracy. In this study, SP and DP values can also be measured directly without any calibration in blood pressure estimation. © 2018 IEEE.Öğe Diagnosis and classification of Parkinson's disease using ensemble learning and 1D-PDCovNN(Pergamon-Elsevier Science Ltd, 2023) Nour, Majid; Şentürk, Ümit; Polat, KemalIn this paper, we proposed a novel approach to diagnose and classify Parkinson's Disease (PD) using ensemble learning and 1D-PDCovNN, a novel deep learning technique. PD is a neurodegenerative disorder; early detection and correct classification are essential for better disease management. The primary aim of this study is to develop a robust approach to diagnosing and classifying PD using EEG signals. As the dataset, we have used the San Diego Resting State EEG dataset to evaluate our proposed method. The proposed method mainly consists of three stages. In the first stage, the Independent Component Analysis (ICA) method has been used as the pre-processing method to filter out the blink noises from the EEG signals. Also, the effect of the band showing motor cortex activity in the 7-30 Hz frequency band of EEG signals in diagnosing and classifying Parkinson's disease from EEG signals has been investigated. In the second stage, the Common Spatial Pattern (CSP) method has been used as the feature extraction to extract useful information from EEG signals. Finally, an ensemble learning approach, Dynamic Classifier Selection (DCS) in Modified Local Accuracy (MLA), has been employed in the third stage, consisting of seven different classifiers. As the classifier method, DCS in MLA, XGBoost, and 1D-PDCovNN classifier has been used to classify the EEG signals as the PD and healthy control (HC). We first used dynamic classifier selection to diagnose and classify Parkinson's disease (PD) from EEG signals, and promising results have been obtained. The performance of the proposed approach has been evaluated using the classification accuracy, F-1 score, kappa score, Jaccard score, ROC curve, recall, and precision values in the classification of PD with the proposed models. In the classification of PD, the combination of DCS in MLA achieved an accuracy of 99,31%. The results of this study demonstrate that the proposed approach can be used as a reliable tool for early diagnosis and classification of PD.Öğe Diagnosis of Alzheimer's disease using boosting classification algorithms(Institute of Electrical and Electronics Engineers Inc., 2023) Önder, Mithat; Şentürk, Ümit; Polat, Kemal; Paulraj, D.Alzheimer's Disease (AD) is a progressive degenerative disorder of the brain that impacts memory, cognition, and, ultimately, the ability to carry out daily activities. There is presently no cure for Alzheimer's Disease. However, there are available treatments to manage symptoms and slow their advancement. This research conducted a comprehensive study to diagnose AD using four different categorization methods. These methods included XGBoost, GradientBoost, AdaBoost, and voting classification algorithms. To carry out the examination, a high-quality dataset was obtained from the collection of machine learning data of the prestigious University of California. This dataset was carefully selected to ensure accurate and reliable results. The analysis of the collected data revealed some interesting findings. XGBoost exhibited an accuracy rate of 85% in diagnosing Alzheimer's Disease. ADABoost also performed, achieving an accuracy rate of 75%. GradientBoost, similarly, obtained an accuracy rate of 85%. Additionally, the voting classification algorithms showed promise, attaining an accuracy rate of 80%. All these accuracy rates were obtained by implementing a 5-fold cross-validation methodology, which ensured robust and unbiased results. This research contributes to the field of AD diagnosis by providing insights into the effectiveness of different categorization methods. © 2023 IEEE.Öğe Electricity consumption forecast using machine learning regression models in Turkey(Institute of Electrical and Electronics Engineers Inc., 2022) Şentürk, Ümit; Beken, Murat; Eyecioğlu, ÖnderIn today's energy crisis, countries need to know their energy consumption and make their energy investments accordingly. The variability of end users demanding energy makes it difficult to estimate energy needs. In this article, it has been tried to forecast the future consumption from the electrical energy data consumed in Turkey between the years 2016-2022. After the electricity consumption data was converted into daily data, electrical energy consumption estimations were made with machine learning methods such as linear regression, tree regression, voting regression, XGB regression and Artificial neural network (ANN) methods. Estimation results were evaluated with Mean Square Error (MSE) and R2 (coefficient of determination) performance metrics. As a result of the evaluations made with the test data, MSE=0.006 (0-1 min-max normalization dataset) and R2= 82.7 performances, voting regression obtained the best result among the methods used. Accurate estimation of energy consumption will enable energy production to be made at the optimum level. © 2022 IEEE.Öğe A non-invasive continuous cuffless blood pressure estimation using dynamic recurrent neural networks(Elsevier Sci Ltd, 2020) Şentürk, Ümit; Polat, Kemal; Yücedağ, İbrahimCardiovascular diseases (CVD) have become the most important health problem of our time. High blood pressure, which is cardiovascular disease, is a risk factor for death, stroke, and heart attack. Blood pressure measurement is commonly used to limit blood flow in the arm or wrist, with the cuff. Since blood pressure cannot be measured continuously in this method, the dynamics underlying blood pressure cannot be determined and are inefficient in capturing symptoms. This paper aims to perform blood pressure estimation using Photoplethysmography (PPG) and Electrocardiography (ECG) signals that do not obstruct the vascular access. These signals were filtered and segmented synchronously from the R interval of the ECG signal, and chaotic, time, and frequency domain features were subtracted, and estimation methods were applied. Different methods of machine learning in blood pressure estimation are compared. Dynamic learning methods such as Recurrent Neural Network (RNN), Nonlinear Autoregressive Network with Exogenous Inputs Neural Networks NARX-NN and Long-Short Term Memory Neural Network (LSTM-NN) used. Estimation results have been evaluated with performance criteria. Systolic Blood Pressure (SBP) error mean +/- standard deviation = 0.0224 +/- (2.211), Diastolic Blood Pressure (DBP) error mean +/- standard deviation = 0.0417 +/- (1.2193) values have been detected in NARX artificial neural network. The blood pressure estimation results are evaluated by the British Hypertension Society (BHS) and American National Standard for Medical Instrumentation ANSI/AAMI SP10: 2002. Finding the most accurate and easy method in blood pressure measurement will contribute to minimizing the errors. (C) 2020 Elsevier Ltd. All rights reserved.Öğe A novel blood pressure estimation method with the combination of long short term memory neural network and principal component analysis based on PPG signals(Springer, 2020) Şentürk, Ümit; Polat, Kemal; Yücedağ, İbrahimThe worldwide high blood pressure-related mortality rate is increasing. Alternative measurement methods are required for blood pressure measurement. There are similarities between blood pressure and photoplethysmography (PPG) signals. In this study, a novel blood pressure estimation methods based on the feature extracted from the PPG signals have been proposed. First of all, 12-time domain features have extracted from the raw PPG signal. Secondly, raw PPG signals have been applied to Principal Component Analysis (PCA) to obtain 10 new features. The resulting features have been combined to form a hybrid feature set consisting of 22 features. After features extraction, blood pressure values have automatically been predicted by using the Long Short Term Memory Neural Network (LSTM-NN) model. The prediction performance measures including MAE, MAPE, RMSE, and IA values have been used. While the combination of 12-time domain features from PPG signals and LSTM has obtained the MAPE values of 0,0547 in the prediction of blood pressures, the combination of 10-PCA coefficients and LSTM has achieved the MAPE value of 0,0559. The combination model of all features (22) and LSTM has obtained the MAPE values of 0,0488 in the prediction of blood pressures. The achieved results have shown that the proposed hybrid model based on combining PCA and LSTM is very promising in the prediction of blood pressure from PPG signals. In the future, the proposed hybrid method can be used as a wearable device in the measurement of blood pressure without any calibration.Öğe A novel cuffless blood pressure prediction: Uncovering new features and new hybrid ml models(MDPI, 2023) Nour, Majid; Polat, Kemal; Şentürk, Ümit; Arıcan, MuratThis paper investigates new feature extraction and regression methods for predicting cuffless blood pressure from PPG signals. Cuffless blood pressure is a technology that measures blood pressure without needing a cuff. This technology can be used in various medical applications, including home health monitoring, clinical uses, and portable devices. The new feature extraction method involves extracting meaningful features (time and chaotic features) from the PPG signals in the prediction of systolic blood pressure (SBP) and diastolic blood pressure (DBP) values. These extracted features are then used as inputs to regression models, which are used to predict cuffless blood pressure. The regression model performances were evaluated using root mean squared error (RMSE), R-2, mean square error (MSE), and the mean absolute error (MAE). The obtained RMSE was 4.277 for systolic blood pressure (SBP) values using the Matern 5/2 Gaussian process regression model. The obtained RMSE was 2.303 for diastolic blood pressure (DBP) values using the rational quadratic Gaussian process regression model. The results of this study have shown that the proposed feature extraction and regression models can predict cuffless blood pressure with reasonable accuracy. This study provides a novel approach for predicting cuffless blood pressure and can be used to develop more accurate models in the future.Öğe A novel IoT-based arrhythmia detection system with ECG signals using a hybrid convolutional neural network and neural architecture search network(Inderscience Publishers, 2024) Şentürk, Ümit; Gülen, Ceren; Polat, KemalElectrocardiogram (ECG) signals are the most common tool to evaluate the heart’s function in cardiovascular diagnosis. Irregular heartbeats (arrhythmia) found in the ECG play an essential role in diagnosing cardiovascular diseases (CVD). In this paper, we proposed an arrhythmia classification method with the neural architecture search network (NASNet) model, an optimised version of the convolutional neural networks (2D-CNN) model, which performs very well in visual information analysis classification. We aimed to use the arrhythmia classification problem in IoT and mobile devices with the NASNet model with low parameter numbers and processing capacity without performance loss. 2D input data have been obtained by converting the classified heart rate signals in the datasets into image files. The 2D data obtained have been classified by machine learning, CNN, and NASNet models. As a result of classification, 2D CNN accuracy was 97.51%, and the NASNet model was 96.89% accuracy. As a result of arrhythmia classification, the accuracy rates of the NASNet and the 2D CNN models were close. In conclusion, the proposed IoT-based arrhythmia detection system with ECG signals using a hybrid CNN and NASNet is a promising tool for the early detection of arrhythmias. Furthermore, it could help to reduce the mortality associated with these potentially fatal conditions. © 2024 Inderscience Enterprises Ltd.Öğe A novel ML approach to prediction of breast cancer: Combining of mad normalization, KMC based feature weighting and AdaBoostM1 classifier(Ieee, 2018) Polat, Kemal; Şentürk, ÜmitBreast cancer is the second most common cancer in our country and in the world. In this study, a breast cancer data set was formed from the findings obtained from experiments conducted in the city of Coimbra of Portugal. There are two sets of data (52 data: healthy group, 64 data belong to patient group) and 9 features in the breast cancer data set of 116 data, both healthy and patient. These nine features are: Age, BMI, Glucose, Insulin, HOMA, Leptin, Adiponectin, MCP-1. In the proposed method, a three-step hybrid structure is proposed to detect the presence of breast cancer. In the first step, the dataset was first normalized by the MAD normalization method. In the second step, k-means clustering (KMC) based feature weighting has been used for weighting the normalized data. Finally, the AdaBoostM1 classifier has been used to classify the weighted data set. Only the combination the AdaBoostM1 classifier with MAD normalization method yielded a 75% classification accuracy in the detection of breast cancer, whereas the hybrid approach achieved 91.37% success. These results show that the proposed system could be used safely to detect breast cancer.Öğe Repetitive neural network (RNN) based blood pressure estimation using PPG and ECG signals(Ieee, 2018) Şentürk, Ümit; Yücedağ, İbrahim; Polat, KemalIn this study, a new hybrid prediction model was proposed by combining ECG (Electrocardiography) and PPG (Photoplethysmographic) signals with a repetitive neural network (RNN) structure to estimate blood pressure continuously. The proposed method consists of two steps. In the first step, a total of 22 time-domain attributes were obtained from PPG and ECG signals to estimate systolic blood pressure (SBP) and diastolic blood pressure (DBP) values. In the second phase, these time-domain attributes are set as input to the RNN model and then the blood pressure is estimated. Within the RNN structure, there are two-way long short-term memory BLSTM (Bidirectational Long-Short Term Memory), LSTM and ReLU (Rectified-Linear unit) layers. The bidirectional LSTM layer has been used to remove the negative affects the blood pressure value of past and future effects of nonlinear physiological changes. The LSTM layers has ensured that learning is deep and that mistakes made are reduced. The ReLU layer has been allowed the neural network to quickly reach its conclusion. The same ECG and PPG signals obtained from the database have been cleared from noise and artifacts. And then ECG and PPG signals have been segmented according to peak values of these signals. The results have shown that RMSE (Root Mean Square Error) between the estimated SBP and the measured SBP with RNN model was 3.63 and the RMSE between the estimated DBP and the measured DBP values was 1.48 with RNN model. It has been seen that the used model has a more learning ability. Thanks to the proposed method, a calibration free blood pressure measurement system using PPG and ECG signals, was developed.Öğe Towards wearable blood pressure measurement systems from biosignals: A review(2019) Şentürk, Ümit; Polat, Kemal; Yücedağ, İbrahimBlood pressure is the pressure by the blood to the vein wall. High blood pressure, which is called silent death, is the cause of nearly 13% of mortality all over the world. Blood pressure is not only measured in the medical environment, but the blood pressure measurement is also a need for people in their daily life. Blood pressure estimation systems with low error rates have been developed besides the new technologies and algorithms. Blood pressure measurements are differentiated as invasive blood pressure (IBP) measurement and noninvasive blood pressure (NIBP) measurement methods. Although IBP measurement provides the most accurate results, it cannot be used in daily life because it can only be performed by qualified medical staff with specialized medical equipment. NIBP measurement is based on measuring physiological signals taken from the body and producing results with decision mechanisms. Oscillometric, pulse transit time (PTT), pulse wave velocity, and feature extraction methods are mentioned in the literature as NIBP. In the oscillometric method of the sphygmomanometer, an electrocardiogram is used in PTT methods as a result of the comparison of signals such as electrocardiography, photoplethysmography, ballistocardiography, and seismocardiography. The increase in the human population and worldwide deaths due to the highly elevated blood pressure makes the need for precise measurements and technological devices more clear. Today, wearable technologies and sensors have been frequently used in the health sector. In this review article, the invasive and noninvasive blood pressure methods, including various biosignals, have been investigated and then compared with each other concerning the measurement of comfort and robust estimation.
