Ayhan Yüksel

Medical image segmentation with transform and moment based features and incremental supervised neural network

In this study, a novel incremental supervised neural network (ISNN) is proposed for the segmentation of medical images. Performance of the ISNN is investigated for tissue segmentation in medical images obtained from various imaging modalities. Two feature extraction methods based on transform and moments are comparatively investigated to segment the tissues in medical images. Two-dimensional (2D) continuous wavelet transform (CWT) and the moments of the gray-level histogram (MGH) are computed in order to form the feature vectors of ultrasound (US) bladder and phantom images, X-ray computerized tomography (CT) and magnetic resonance (MR) head images. In the 2D-CWT method, feature vectors are formed by the intensity of one pixel of each wavelet-plane of different energy bands. The MGH represents the tissues within the sub-windows by using the spatial variation of image intensities. In this study, the ISNN and Grow and Learn (GAL) network are employed for the segmentation task. It is observed that the ISNN has significantly eliminated the disadvantages of the GAL network in the segmentation of the medical images.

A Neural Network-Based Optimal Spatial Filter Design Method for Motor Imagery Classification

In this study, a novel spatial filter design method is introduced. Spatial filtering is an important processing step for feature extraction in motor imagery-based brain-computer interfaces. This paper introduces a new motor imagery signal classification method combined with spatial filter optimization. We simultaneously train the spatial filter and the classifier using a neural network approach. The proposed spatial filter network (SFN) is composed of two layers: a spatial filtering layer and a classifier layer. These two layers are linked to each other with non-linear mapping functions. The proposed method addresses two shortcomings of the common spatial patterns (CSP) algorithm. First, CSP aims to maximize the between-classes variance while ignoring the minimization of within-classes variances. Consequently, the features obtained using the CSP method may have large within-classes variances. Second, the maximizing optimization function of CSP increases the classification accuracy indirectly because an independent classifier is used after the CSP method. With SFN, we aimed to maximize the between-classes variance while minimizing within-classes variances and simultaneously optimizing the spatial filter and the classifier. To classify motor imagery EEG signals, we modified the well-known feed-forward structure and derived forward and backward equations that correspond to the proposed structure. We tested our algorithm on simple toy data. Then, we compared the SFN with conventional CSP and its multi-class version, called one-versus-rest CSP, on two data sets from BCI competition III. The evaluation results demonstrate that SFN is a good alternative for classifying motor imagery EEG signals with increased classification accuracy.

Analog filter group delay optimization using the Vortex Search algorithm

In this study, a method based on the Vortex Search algorithm is proposed for analog filter group delay optimization. To measure the performance of the proposed method, a number of all-pass filters are first cascaded to a fifth order Chebyshev low-pass filter and then the optimum parameters of these all-pass filters are searched by using the Vortex Search algorithm. It is shown that, the group delay of the fifth order Chebyshev low-pass filter (2.593s) can be decreased down to 1.547s by using an optimum fourth order all-pass filter.

Simulation of normal cardiovascular system and severe aortic stenosis using equivalent electronic model.

OBJECTIVE In this study, we have designed an analog circuit model of the cardiovascular system that is able to simulate normal condition and cardiovascular diseases, such as mitral stenosis, aortic stenosis, and hypertension. Especially we focused on severe aortic stenosis, because it is one of the causes of sudden death in asymptomatic patients. In this study, we aim to investigate the simulation of the cardiovascular system using an electronic circuit model under normal and especially severe aortic valve stenosis conditions. METHODS The Westkessel model including RLC pi-segments is chosen in order to simulate both systemic and pulmonary circulation. The left and right heart is represented by trapezoidal shape stiffnesses. Aortic capacitance and aortic valve characteristics are chosen nonlinear. Severe aortic stenosis is implemented by changing the value of the serial resistance to the aortic valve. MATLAB software program is used for the model implementation. RESULTS The results for normal conditions of the given electrical model are similar to the normal cardiovascular physiology. As a result of simulation, a remarkable increase of the left ventricle systolic blood pressure and aortic mean pressure gradient, and decrease of aortic systolic blood pressure are observed in severe aortic valve stenosis. CONCLUSION In conclusion, our model is effective and available for simulating normal cardiac conditions and cardiovascular diseases, especially severe aortic stenosis.

Dimension reduction by a novel unified scheme using divergence analysis and genetic search

In this study, a unified scheme using divergence analysis and genetic search is proposed to determine significant components of feature vectors in high-dimensional spaces, without having to deal with singular matrix problems. In the literature it is observed that three main problems exist in the feature selection process performed in a high-dimensional space. These problems are high computational load, local minima, and singular matrices. In this study, feature selection is realized by increasing the dimension one by one, rather than reducing the dimension. In this sense, the recursive covariance matrices are formulated to decrease the computational load. The use of genetic algorithms is proposed to avoid local optima and singular matrix problems in high-dimensional feature spaces. Candidate strings in the genetic pool represent the new features formed by increasing the dimension. The genetic algorithms investigate the combination of features which give the highest divergence value. In this study, two methods are proposed for the selection of features. In the first method, features in a high-dimensional space are determined by using divergence analysis and genetic search (DAGS) together. If the dimension is not high, the second method is offered which uses only recursive divergence analysis (RDA) without any genetic search. In Section 3 two experiments are presented: Feature determination in a two-dimensional phantom feature space, and feature determination for ECG beat classification in a real data space.

Retrospective correction of near field effect of X-ray source in radiographic images by using genetic algorithms

X-ray bone images are used in the areas such as bone age assessment, bone mass assessment and examination of bone fractures. Medical image analysis is a very challenging problem due to large variability in topologies, medical structure complexities and poor image modalities such as noise, low contrast, several kinds of artifacts and restrictive scanning methods. Computer aided analysis leads to operator independent, subjective and fast results. In this study, near field effect of X-ray source is eliminated from hand radiographic images. Firstly, near field effect of X-ray source is modeled, then the parameters of the model are estimated by using genetic algorithms. Near field effect is corrected for all image pixels retrospectively. Two different categories of images are analyzed to show the performance of the developed algorithm. These are original X-ray hand images and phantom hand images. Phantom hand images are used to analyze the effect of noise. Two performance criteria are proposed to test the developed algorithm: Hand segmentation performance and variance value of the pixels in the background. It is observed that the variance value of the pixels in the background decreases, and hand segmentation performance increases after retrospective correction process is applied.

Automatic segmentation of bone tissue in X-Ray hand images

Automatic segmentation of X-ray hand images is an important process. In studies such as skeletal bone age assessment, bone densitometry and analyzing of bone fractures, it is a necessary extremely difficult and complicated task. In this study, hand X-ray images were segmented by using C-means classifier. Extraction of bone tissue was realized in three steps: i) preprocessing, ii) feature extraction and iii) automatic segmentation. In preprocessing scheme, inhomogeneous intensity distribution is eliminated and some structural pre-information about hand was obtained in order to use in feature extraction block. In feature extraction process, edges between soft and bone tissues were extracted by proposed enhancement process. In automatic segmentation process, the image was segmented using C-mean classifier by taking care of local information. In the study, hand images of ten different people were segmented with high performances above 95%.

Filter Bank Common Spatio-Spectral Patterns for Motor Imagery Classification

In this study, a new spatio-spectral filtering method for motor imagery signal analysis is introduced. Motor imagery is an important research area in brain computer interfacing. EEG signals related with motor imagery have characteristic frequencies originating from sensorimotor cortex. Common spatial patterns (CSP) method is a very popular and successful spatial filtering algorithm in motor imagery classification. However, CSP only optimizes spatial filters, subject specific frequency selection should be done manually, which is a meticulous process. Therefore, an automatic method for spectral filter optimization is needed. Proposed filter bank common spatio-spectral patterns (FBCSSP) algorithm optimizes spatial and spectral filters. FBCSSP method uses a network of a filter bank and two consecutive CSP layers so that proposed structure has a subject specific response in both spatial and spectral domains. We inspected the proposed method in terms of classification accuracy and physiological consistence of the created filters using publicly available data set. FBCSSP method gave higher classification accuracy than other spatio-spectral pattern methods in the literature. Also, obtained spatial and spectral filters were consistent with the spatial and spectral properties of motor imagery signals.

Modeling of inhomogeneous intensityd istribution of X-ray source in radiographic images

X-ray bone images are used in the areas such as bone age assessment, bone mass assessment and examination of bone fractures. The computer aided analysis leads to operator independent, subjective and fast results.

Divergent common spatial patterns method

In this study, a novel method called DIVCSP is proposed that can be a good alternative to the spatial filtering methods which are used to have a good discrimination between different classes in motor imagery based brain computer interfaces. With DIVCSP method spatial filters that increase between classes scattering while decreasing in-class scattering can be obtained thus, the classification accuracy can be increased. The classification results obtained with DIVCSP show that, proposed method is a good alternative to CSP method which is widely popular.