Lale Ozyilmaz

Diagnosis of thyroid disease using artificial neural network methods

Proper interpretation of the thyroid gland functional data is an important issue on the diagnosis of thyroid disease. The primary role of the thyroid gland is to help regulation of the bodys metabolism. Thyroid hormone produced by the thyroid gland provides this. Production of too little thyroid hormone (hypo-thyroidism) or production of too much thyroid hormone (hyper-thyroidism) definites the type of thyroid disease. In this work, various neural network methods have been used to help diagnosis of thyroid disease.

Artificial neural networks for diagnosis of hepatitis disease

Recently, neural networks have become a very important method in the field of medical diagnostics. The objective of this work is to diagnose hepatitis disease by using different neural network architectures. Standard feedforward networks and a hybrid network were investigated. Results obtained show that especially the hybrid network can be successfully used for diagnosing of hepatitis.

EMG signal classification using conic section function neural networks

The aim of this work is to classify EMG signals using a new neural network architecture to control multifunction prostheses. The control of these prostheses can be made using myoelectric signals taken from a single pair of surface electrodes. This case has been demonstrated specifically for use by above elbow amputees. The ability to separate different muscle contraction characters depends on myoelectric signal information. Therefore, the classification of these signals is investigated. The proposed neural network algorithm here makes the user learn better and faster.

Dimensionality reduction in conic section function neural network

This paper details how dimensionality can be reduced in conic section function neural networks (CSFNN). This is particularly important for hardware implementation of networks. One of the main problems to be solved when considering the hardware design is the high connectivity requirement. If the effect that each of the network inputs has on the network output after training a neural network is known, then some inputs can be removed from the network. Consequently, the dimensionality of the network, and hence, the connectivity and the training time can be reduced. Sensitivity analysis, which extracts the cause and effect relationship between the inputs and outputs of the network, has been proposed as a method to achieve this and is investigated for Iris plant, thyroid disease and ionosphere databases. Simulations demonstrate the validity of the method used.

Modelling of active microwave transistors using artificial neural networks

Signal and noise behaviours of microwave transistors are modeled through the neural network approach for the whole operating ranges including frequency bias and configuration types. Here, the device is modeled by a black box whose small-signal and noise parameters are evaluated through various neural network methods, based upon the fitting of both of these parameters for multiple bias and configuration. Previous results are improved with a conic section function neural network method presented in this work.

ROC analysis for fetal hypoxia problem by artificial neural networks

As fetal hypoxia may damage or kill the fetus, it is very important to monitor the infant so that any signs of fetal distress can be detected as soon as possible. In this paper, the performances of some artificial neural networks are evaluated, which eventually produce the suggested diagnosis of fetal hypoxia. Multilayer perceptron (MLP) structure with standard back propagation, MLP with fast back propagation (adaptive learning and momentum term added), Radial Basis Function (RBF) network structure trained by orthogonal least square algorithm, and Conic Section Function Neural Network (CSFNN) with adaptive learning were used for this purpose. Further more, Receiver Operating Characteristic (ROC) analysis is used to determine the accuracy of diagnostic test.

Sensitivity analysis for conic section function neural networks

Sensitivity analysis is a method for extracting the cause and effect relationship between the inputs and outputs of the network. After training a neural network, one may want to know the effect that each of the network inputs is having on the network output. The basic idea is that each input channel to the network is offset slightly and the corresponding change in the output(s) is reported. The input channels that produce low sensitivity values can be considered insignificant and can most often be removed from the network. This will reduce the size of the network, which in turn reduces the complexity and the training time. Furthermore, this may also improve the network performance. In this work, sensitivity analysis for conic section function neural network is investigated and the results are given for different problems.

A computational approach to estimate postmortem interval using opacity development of eye for human subjects

This paper presents an approach to postmortem interval (PMI) estimation, which is a very debated and complicated area of forensic science. Most of the reported methods to determine PMI in the literature are not practical because of the need for skilled persons and significant amounts of time, and give unsatisfactory results. Additionally, the error margin of PMI estimation increases proportionally with elapsed time after death. It is crucial to develop practical PMI estimation methods for forensic science. In this study, a computational system is developed to determine the PMI of human subjects by investigating postmortem opacity development of the eye. Relevant features from the eye images were extracted using image processing techniques to reflect gradual opacity development. The features were then investigated to predict the time after death using machine learning methods. The experimental results prove that the development of opacity can be utilized as a practical computational tool to determine PMI for human subjects.

Character/text data compression and encryption by compressive sensing and hybrid cryptography

Compressed sensing (CS) is a methodology allowing linear measurements much fewer in number than the length of the original signal vectors. This is made possible by using a measurement matrix which converts the original signal to a much shorter signal. For CS to work, it is necessary that the original signal is sufficiently sparse so that it can be reconstructed from the compressed signal. Hybrid cryptography combines symmetric and asymmetric cryptographies. In order to increase security, the symmetric keys used are transmitted to the receiver by asymmetric cryptography. In this study, the proposed character/text input is proposed to be sensed by compressive sensing using the method of orthogonal matching pursuit (OMP), and then to be encrypted by hybrid cryptography using a transform and amplitude-phase keys similarly to a 4f system of optical cryptography. The overall system achieves both data compression and encryption.

Cryptography with compressive sensing Orthogonal Matching Pursuit method

Compressive sensing is a new technique of reconstructing a signal with much less number of measurements than what is required by the sampling theorem. Sparse signals, detection (measurement) matrices are obtained with the aid of measurements stored into vectors. If the signal is sparse enough, the original signal can be recovered successfully. In this study, compressive sensing Orthogonal Matching Pursuit (OMP) algorithm is used, and applied to symmetric cryptography application. Since OMP is a compressive sensing method, both compressive sensing and cryptographic encyrption/decyrption are achieved simultaneously. Two-times complex reinforced encryption and decryption cryptography was also achieved with the new method.