الفهرس 
Chapter One IntroductionOnly 14 pages are availabe for public view
 |  | from | 228 |  |  |
| | | from | 228 | | |
Abstract
A wide range of signals is obtained from the human body, known as Biomedical signals or biosignals, they can be at the cell position, organ position, or sub-atomic position. electrical activity from the muscle sound signals appertained to as electromyogram, the electrical exertion from the encephalon known as an electroencephalogram, the electrical activity from the heart called an electrocardiogram, the electroretinogram from the eye, and so on. Studying and improving this field of research is very important to physicians whose work is related to this branch of medicine, monitoring and observing changes in these signals helps them to cover, predict, and cure various diseases, it can help them examine and prognosticate and cure numerous conditions. Still, these signals are frequently affected by the collection of various types of noise, these various types of noise that contaminate medical signals are power line interference, electrode contact noise, motion artifacts, muscle contraction, base line wander, instrumentation noise generated by electronic devices and electrosurgical noise, it’s important to remove this noise from the signals to get useful information. Noise removal is complicated due to the time-varying nature of medical signals, throw years scientists and researchers have developed many techniques for noise removal, they tried various approaches based on wavelet transform, Fuzzy logic, finite impulse response filtering, empirical mode decomposition, and blind source separation used in denoising the signal effectively. In this work we have used in this thesis the well-known blind source separation algorithm called independent components analysis for medical signals denoising, the noise removal process is done by proposing new flexible score functions families for blind source separation, using the generalized Gamma densities family, the exponentiated transmuted Weibull densities family, the fractional Weibull densities family, and the modified Weibull densities family. To blindly get the independent source signals, we apply the well-known Fast Independent Component Analysis (Fast-ICA) algorithm and the statistically principled method called sparse code shrinkage, the parameter estimation of similar score functions is achieved by using an effective system based on maximum likelihood. The results attained using the proposed densities in our mechanisms are better than those attained by other distribution functions. This thesis is organized into six chapters, in the first chapter we introduce the blind source separation problem and the independent component analysis and their importance in the field of source separation and noise removal. In the second chapter, we discuss the generalized Gamma distribution model for noise removal from both medical signals and medical images. Chapter three discusses the biomedical signals denoising based on exponentiated transmuted Weibull distribution. Chapter four uses fractional Weibull distribution for biomedical signals denoising and biomedical images denoising. Chapter five shows the usage of modified Weibull distribution for the enhancement of both medical signals and images. Chapter six, we give the main results of the thesis and it’s interesting to be applied. Moreover, we discuss future work and new results that may be able to extend our present work.