Kaya, Ceren | Erkaymaz, Okan | Ayar, Orhan | Özer, Mahmut

Proceedings | 2018 | 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 , pp.1 - 4

Refractive disorders are common health problems in the community and they are the most important cause of visual impairment. In this study, it was aimed to classify the individuals who have hypermetropia and myopia refractive disorders or not. For this, horizontal and vertical Electrooculogram (EOG) signal data from the right and left eyes of the individuals were used. The performance of the data was investigated by using Logistic Regression (LR), Naive Bayes (NB), Random Forest (RF) and REP Tree (RT) data mining methods. According to the obtained results, REP Tree method has shown the most successful classification performance to d . . .etect hypermetropia and myopia refractive disorders from Electrooculogram (EOG) signals. © 2018 IEEE Daha fazlası Daha az

Kaya, Ceren | Erkaymaz, Okan | Ayar, Orhan | Özer, Mahmut

Proceedings | 2017 | 2017 Medical Technologies National Conference, TIPTEKNO 20172017-January , pp.1 - 4

'Diabetes Mellitus (Diabetes)' is a disease based on insulin hormone disorders secreted from the pancreas gland. Clinical findings find out that diabetes causes some diseases in vital organs. 'Diabetic Retinopathy' is one of the most common eye diseases based on diabetes, and it is the leading cause of visual loss resulting from structural changes in the retinal vessels. Recent researches show that signals from vital organs can be used to diagnose diseases in the literature. In this study, the features of horizontal and vertical Video-Oculography (VOG) signals from right and left eye are used to classify non-proliferative and prolif . . .erative diabetic retinopathy disease. 25 statistical features are obtained using discrete wavelet transform with VOG signals from 24 subjects. Feature selection is performed using C4.5 decision tree algorithm from 25 features obtained. The statistical features obtained from C4.5 decision tree and discrete wavelet transform are applied as input to artificial neural networks and the classification performance of the 'Diabetic Retinopathy' disease are compared according to these two methods. Our results show that feature selection by C4.5 decision tree algorithm (96.87%) provides better classification performance than feature extraction with discrete wavelet transform (93.75%). © 2017 IEEE Daha fazlası Daha az

Erkaymaz, Hande | Özer, Mahmut | Kaya, Ceren

Conference Object | 2015 | 2015 19TH NATIONAL BIOMEDICAL ENGINEERING MEETING (BIYOMUT) , pp.1 - 4

The eye, which has the most advanced features among sense organs, has a perfect functioning on people. Furthermore, as it is placed in the lead role of vision, shows the importance for people is quite big. Nowadays, biomedical devices developed for patients who have mobility are benefiting from eye movements. Electrooculogram studies are especially designed on the basis of the signal depending on the movement of your eyes. Electrical origin of EOG biological signal, that occur around the eye pupil, makes an attempt to meet the needs of patients by the right, left, up, down and blinking action. In this study, 4 basic differences exis . . .ting in the direction of movement using voltage controlled EOG signal studies have tried to determine the Neuro-Fuzzy model. Determining the direction of Neuro-Fuzzy control system demonstrates how it can be successfully used as. In addition, control algorithms of artificial intelligence systems that use this kind of eye signals benefiting from the input of the detection process is advantageous in the classification of complex environment Daha fazlası Daha az

Aydemir, Zübeyr | Erkaymaz, Okan | Ferah, Meryem Akpolat

Proceedings | 2017 | 2017 25th Signal Processing and Communications Applications Conference, SIU 2017 , pp.1 - 4

In this study, the features of the seminiferous tubule sections were extracted and the presence of the cells and cell stain types detected with the help of the feed forward artificial neural network. By looking at the section view with a small window, 78 features were extracted from the pixels seen by the window and used as an input to the artificial neural network. Artificial neural network outputs are decides presence of the cell and the staining of the cell. The results obtained with the artificial neural network were determined by using the connected component labeling method. The results obtained with the help of the user and t . . .he results obtained with the artificial neural network were compared. It has been shown that the proposed ANN model performs cell counting process comparable to the literature (%76 accuracy). © 2017 IEEE Daha fazlası Daha az

Erkaymaz, Okan | Palabaş, Tuğba

Proceedings | 2018 | 26th IEEE Signal Processing and Communications Applications Conference, SIU 2018 , pp.1 - 4

Computer assisted automatic diagnostic systems are used for the purpose of speeding up diagnosis and treatment and helping to make the right decision. In this study, cervical cancer is identified using four basic classifiers: Naive Bayes (NB), k-Nearest Neighbor (kNN), Multilayer Perceptron (MLP) and Decision Trees (KA-C4.5) algorithms and random subspaces ensemble algorithm. Gain Ratio Attribute Evaluation (GRAE) feature extraction algorithm is applied to contribute to classification performance. The classification results obtained with all datasets and reduced datasets are compared with respect to performance criteria such as accu . . .racy, Root Mean Square Error (RMSE), Sensitivity, Specificity performance criteria. According to the obtained performance analysis, it is seen that the classification performance with the random subspace ensemble algorithm using the kNN basic classifier on the reduced data set is the highest (%95.51). © 2018 IEEE Daha fazlası Daha az

Kara, Ferdi | Kaya, Hakan | Erkaymaz, Okan | Öztürk, Ertan

Proceedings | 2016 | Proceedings of the 2016 International Symposium on INnovations in Intelligent SysTems and Applications, INISTA 2016 , pp.1 - 4

In wireless communications, the cooperative communication (CC) technology promises performance gains compared to traditional Single-Input Single Output (SISO) techniques. Therefore, the CC technique is one of the nominees for 5G networks. In the Decode-And-Forward (DF) relaying scheme which is one of the CC techniques, determination of the threshold value at the relay has a key role for the system performance and power usage. In this paper, we propose prediction of the optimal threshold values for the best relay selection scheme in cooperative communications, based on Artificial Neural Networks (ANNs) for the first time in literatur . . .e. The average link qualities and number of relays have been used as inputs in the prediction of optimal threshold values using Artificial Neural Networks (ANNs): Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF) networks. The MLP network has better performance from the RBF network on the prediction of optimal threshold value when the same number of neurons is used at the hidden layer for both networks. Besides, the optimal threshold values obtained using ANNs are verified by the optimal threshold values obtained numerically using the closed form expression derived for the system. The results show that the optimal threshold values obtained by ANNs on the best relay selection scheme provide a minimum Bit-Error-Rate (BER) because of the reduction of the probability that error propagation may occur. Also, for the same BER performance goal, prediction of optimal threshold values provides 2dB less power usage, which is great gain in terms of green communication. © 2016 IEEE Daha fazlası Daha az

Erkaymaz, Hande | Özer, Mahmut | Kaya, Ceren | Orak, I. Muharrem

Proceedings | 2015 | 2015 23RD SIGNAL PROCESSING AND COMMUNICATIONS APPLICATIONS CONFERENCE (SIU) , pp.1228 - 1231

Nowadays, artificial movements have been obtained by utilizing other organs for paralyzed patients. Especially the usage of eye movements for giving message to outside world became popular as a scientific subject. In studies according to eye movements, the Electrooculogram (EOG) signal is used. In this study, the vertical and horizontal FOG signals taken from electrodes, placed around the eyes, have been modelled by using Artificial Neural Networks (ANN) which is one of artificial intelligent technique. The system can sense four main directions (Right, Left, Up and Down) at the same time it can also detect blinking movements. Firstl . . .y, the signals have been pre-filtered, amplified and classified by ANN. The performance of recommended model has been demonstrated by analyzing statistical accuracy and confusion matrix according to the features of obtained signal. It has been seen that eye movements can be successfully determined by designed model Daha fazlası Daha az

Kaya, Ceren | Erkaymaz, Okan | Ayar, Orhan | Özer, Mahmut

Proceedings | 2017 | 2016 Medical Technologies National Conference, TIPTEKNO 2016 , pp.1228 - 1231

The insulin hormone secreted from the pancreas gland in the body is not present in sufficient amount, or because they do not fit, which is defined as the elevation of blood glucose 'Diabetes Mellitus (Diabetes)'. 'Diabetic Retinopathy' is the most common in diabetes-related eye diseases. It had done damages in the retina that detect light on behind the eye as a result of changes in the arteries that is one of the reasons that makes blindness (loss of vision) in people. In this study, horizontal and vertical Video-Oculography (VOG) signals captured by using internal tracking camera in Metrovision MonPackOne Electrooculography device. . . . In order to filter the noise from the signals, the wavelet transform method was used. Obtained signals have shown that the signals of diabetic retinopathy patients have higher amplitude and irregular characteristic than the signals obtained from healthy groups. In both groups, significant Daubechies-6 wavelet coefficients (A6-D6) gave better results than Daubechies-4 wavelet coefficients (A4-D4). Obtained data as a result of using wavelet transform sheds light on feature extraction and classification in proposed future works. © 2016 IEEE Daha fazlası Daha az

Kaya, Ceren | Erkaymaz, Okan | Ayar, Orhan | Özer, Mahmut

Proceedings | 2017 | 2017 25th Signal Processing and Communications Applications Conference, SIU 2017 , pp.1228 - 1231

25th Signal Processing and Communications Applications Conference, SIU 2017 -- 15 May 2017 through 18 May 2017 -- -- 128703

Erkaymaz, Okan | Senyer, İrem | Uzun, Rukiye

Proceedings | 2017 | 2017 25th Signal Processing and Communications Applications Conference, SIU 2017 , pp.1228 - 1231

Using surface EMG signals is a non-invasive measurement method obtained as a result of muscle activity. In this study, surface EMG data have been used for classification, taken from healthy individuals or individuals with knee abnormalities in gait position. For this purpose, first feature extraction was realized by discrete wavelet transform from the data. Then, extracted features were classified by artificial neural network approach that is widely used in the literature. In classification process, artificial neural networks were trained by using simple cross-validation algorithm. During training the optimal network topology was de . . .termined. The highest classification performance of proposed model was obtained in rate fiction 80%-20% and 70%-30% of data set. Our results revealed that proposed artificial neural network model is able to detect knee abnormality from surface EMG signals. © 2017 IEEE Daha fazlası Daha az