Baha Sen

A Novel Method for Automated Diagnosis of Epilepsy Using Complex-Valued Classifiers

The study reported herein proposes a new method for the diagnosis of epilepsy from electroencephalography (EEG) signals based on complex classifiers. To carry out this study, first the features of EEG data are extracted using a dual-tree complex wavelet transformation at different levels of granularity to obtain size reduction. In subsequent phases, five features (based on statistical measurements maximum value, minimum value, arithmetic mean, standard deviation, median value) are obtained by using the feature vectors, and are presented as the input dimension to the complex-valued neural networks. The evaluation of the proposed method is conducted using the k-fold cross-validation methodology, reporting on classification accuracy, sensitivity, and specificity. The proposed method is tested using a benchmark EEG dataset, and high accuracy rates were obtained. The stated results show that the proposed method can be used to design an accurate classification system for epilepsy diagnosis.

Computer-Aided Diagnosis of Parkinson's Disease Using Complex-Valued Neural Networks and mRMR Feature Selection Algorithm.

Parkinsons disease (PD) is a neurological disorder which has a significant social and economic impact. PD is diagnosed by clinical observation and evaluations, coupled with a PD rating scale. However, these methods may be insufficient, especially in the initial phase of the disease. The processes are tedious and time-consuming, and hence systems that can automatically offer a diagnosis are needed. In this study, a novel method for the diagnosis of PD is proposed. Biomedical sound measurements obtained from continuous phonation samples were used as attributes. First, a minimum redundancy maximum relevance (mRMR) attribute selection algorithm was applied for the identification of the effective attributes. After conversion to a complex number, the resulting attributes are presented as input data to the complex-valued artificial neural network (CVANN). The proposed novel system might be a powerful tool for effective diagnosis of PD.

Sparsity-Driven Despeckling for SAR Images

Speckle noise inherent in synthetic aperture radar (SAR) images seriously affects the result of various SAR image processing tasks such as edge detection and segmentation. Thus, speckle reduction is critical and is used as a preprocessing step for smoothing homogeneous regions while preserving features such as edges and point scatterers. Although state-of-the-art methods provide better despeckling compared with conventional methods, their resource consumption is higher. In this letter, a sparsity-driven total-variation (TV) approach employing l0-norm, fractional norm, or l1-norm to smooth homogeneous regions with minimal degradation in edges and point scatterers is proposed. Proposed method, sparsity-driven despeckling (SDD), is capable of using different norms controlled by a single parameter and provides better or similar despeckling compared with the state-of-the-art methods with shorter execution times. Despeckling performance and execution time of the SDD are shown using synthetic and real-world SAR images.

A Parallel Simulated Annealing Algorithm for Weapon-Target Assignment Problem

Weapon-target assignment (WTA) is a combinatorial optimization problem and is known to be NP-complete. The WTA aims to best assignment of weapons to targets to minimize the total expected value of the surviving targets. Exact methods can solve only small-size problems in a reasonable time. Although many heuristic methods have been studied for the WTA in the literature, a few parallel methods have been proposed. This paper presents parallel simulated algorithm (PSA) to solve the WTA. The PSA runs on GPU using CUDA platform. Multi-start technique is used in PSA to improve quality of solutions. 12 problem instances (up to 200 weapons and 200 targets) generated randomly are used to test the effectiveness of the PSA. Computational experiments show that the PSA outperforms SA on average and runs up to 250x faster than a single-core CPU.

A novel hybrid method for determining the depth of anesthesia level: Combining ReliefF feature selection and random forest algorithm (ReliefF+RF)

Depth of anesthesia is a matter of great importance in surgery. Determination of depth of anesthesia is a time consuming and difficult task carried out by experts. This study aims to decide a method that can classify EEG data automatically with a high accuracy and, so will help the experts for determination process. This study consists of three stages: feature extraction of EEG signals, feature selection, and classification. In the feature extraction stage, 41 feature parameters are obtained. Feature selection stage is important to eliminate redundant attributes and improve prediction accuracy and performance in terms of computational time. Effective feature selection algorithms such as minimum redundancy maximum relevance (mRMR); ReliefF; and Sequential Forward Selection (SFS) are preferred at the feature selection stage to select a set of features which best represent EEG signals. These obtained features are used as input parameters of the classification algorithms. At the classification stage, six different classification algorithms such as random forest (RF); feed-forward neural network (FFNN); C4.5 decision tree algorithm (C4.5); support vector machines (SVM); naive bayes; and radial basis function neural network (RBF) are preferred to classify the problem. A comparison is provided between computation times and accuracy rates of these different classification algorithms. The experimental results show that better results according to other classifiers when the obtained attributes by ReliefF algorithm are used with RF classifier.

Automatic segmentation of region of interest for dry eye disease diagnosis system

Today, dry eye disease is a widely seen health problem. It is known that the disease affects %25-%30 of society. In the absence of early diagnosis and treatment, it may cause the occurrence of significant ocular surface damage and then formation of severe eye discomforts. Because of this, at the point of diagnosis and treatment of the common disease an automatic system that provides faster, more accurate and objective decision-making process rather than clinical care is needed to facilitate the work of the experts. Accurate determination of region of interest has high priority for the algorithm to be applied in the diagnosis of dry eye disease. In this study, automatic detection and extraction of region of interest is studied on real dry eye patient data received after applying clinical fluorescein staining test.

Detection of non-coding RNA's with optimized support vector machines

Non-coding RNAs (ncRNAs) are started to work by a lot of scientists in recent years. ncRNAs are playing important roles in the cell and many of them are waiting to be discovered. The Support Vector Machine (SVM) is quite widely used machine learning algorithm in classification problems. Classification process is being difficult when number of problem instances is increased. The classication processes that will take a lot of time when executed on CPU, can be run and optimized in parallel by using multi core platform which is provided by GPU. In this study, detection of ncRNAs was studied by using a large scaled genomic sequence dataset. NVIDIA CUDA parallel programming technology is utilized to be able to accelerate training and test processes that were implemented by SVM. At the end of this study, detection of ncRNAs by using GPU is successfully implemented in shorter time than CPU and with the same success.

A comparison of different classification algorithms for determining the depth of anesthesia level on a new set of attributes

The effect of anesthesia on patient is expressed as the depth of anesthesia. The detection of appropriate depth of anesthesia is a matter of great importance in surgery. Too deep or too little anesthesia implementation may lead to many psychological and physical disorders on patients. Therefore it is necessary to keep the patient at the most appropriate level of anesthesia. This process is important and challenging operation. In this study, a system is proposed which can be used to determine the depth of anesthesia in order to assist physician. Anesthetic substances significantly affect the cortex of the brain. There are studies for determination of depth of anesthesia by monitoring of brain activity. In this study, EEG signals that reflect the brain activity are utilized to measure the depth of anesthesia. The study consists of feature extraction and classification stages of the EEG signal. In the feature extraction stage, a new attribute set consisting of 44 attributes in different categories was obtained. In this way, it is aimed to create an effective set of attributes that can represent EEG signals. The obtained attributes were used as input parameters for classification algorithms. In classification stage, the classification problem is classified by seven different classification algorithms. In this way, comparison of calculation time and accuracy for obtained results in different classification algorithms was provided. With the proposed method for the determination of different depth of anesthesia, 98.169% classification accuracy was achieved.

Fast feature preserving despeckling

Synthetic Aperture Radar (SAR) images contain high amount of speckle noise which causes edge detection, shape analysis, classification, segmentation, change detection and target recognition tasks become more difficult. To overcome such difficulties, smoothing of homogenous regions while preserving point scatterers and edges during speckle reduction is quite important. Besides, due to huge size of SAR images in remote sensing applications efficiency of computational load and memory consumption must be further improved. In this paper, a parallel computational approach is proposed for the Feature Preserving Despeckling (FPD) method which is chosen due to its success in speckle reduction. Speckle reduction performance, execution time and memory consumption of the proposed Fast FPD (FFPD) method is shown using spot mode SAR images.

Scattered data estimation on medical images for cranioplasty applications

Cranioplasty is a surgical operation to repair hole or defects on skull. 3 dimensional computed tomography (CT) images are used for automatic determination of the shape of implant which is used for repairing defect. The designing implant by mathematical model and manufacturing it before operation lowers the operation cost. In this paper, previous studies are examined, applications are realized by radial basis functions (RBF) and insufficient sections of previous studies are revealed.