An Electrocardiogram (ECG) records electrical signals from the heart to detect abnormal heart rhythms or cardiac arrhythmias. Atrial Fibrillation (AF) is the most common arrhythmia which leads to a large number of deaths annually. The diagnosis of heart disease is skill-dependent and time-consuming, therefore using an intelligent system is a time- and cost-effective approach which can also enhance diagnostic accuracy. This study uses several types of Neural Networks (NNs) including the Deep Neural Network (DNN) GoogLeNet, Multi-Layer Perceptron (MLP), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Long Short-Term Memory (LSTM) to identify arrhythmias in AF signals. The results obtained are compared in order to identify the most effective and accurate system for AF diagnosis. The proposed system has two main steps, preprocessing and postprocessing. In the preprocessing step, different approaches based on the classifier network are used. More specifically, for MLP, ANFIS, and LSTM the 1-D Daubechies wavelet is used, and the extracted wavelet coefficients and statistical features are used as input data to the network. For GoogLeNet, the Continuous Wavelet Transform (CWT) is used to create a time-frequency representation of the signal (scalogram) and extract key signal features. In the postprocessing step, the data obtained (extracted features) are used as the input data to classify the signals. Also, the train and test accuracies and the running times are compared. The results obtained indicate that GoogLeNet provides the best accuracy, but its running time is long. Further, although the ANFIS and MLP networks are much faster than LSTM and GoogLeNet, their accuracy is much lower. / Graduate

This diploma thesis deals with the topic of the ECG quality evaluation. The theoretical part of the thesis contains an overview of the methods, which were studied and an explanation of the basic principles connected with the quality evaluation of the ECG signals. The practical part deals with the implementation of three selected methods, one of which is the continuous evaluation of signal quality by means of SNR (signal to noise ratio) calculation. The results of these methods are further discussed and compared.

The thesis focuses on the use of wavelet wiener filtration to remove muscular interference from ECG signals. As part of it, a filter has been implemented in the MATLAB programming environment. The main part of the thesis deals with the optimization of numerical parameters of the proposed filter. The results of the filtration are compared with the results reported by other authors.

This master thesis deals with QRS complex detection and ECG delineation. The theoretical part of this work describes wavelet transform, some of QRS detection approaches and some of ECG delineation approaches. For algorithm realization in Matlab is used redundant dyadic discrete wavelet transform. Algorithm is designed for experimental electrocardiograms of isolated rabbit hearts and it is evaluated through manually determined references.

This dissertation deals with possibilities of using wavelet transforms for elimination of broadband muscle noise in ECG signals. In this work, the characteristics of ECG signals and particularly the most frequently occurring type of interference are discussed firstly. The theory of wavelet transforms is also introduced and followed by design of the simple wavelet filter and the more sophisticated version with wiener filtering of wavelet coefficients. Next part is devoted to the design of our filter, which is based on wavelet wiener filtering and is complemented by algorithms that ensure full adaptability of its parameters when the properties of the input signal are changing. Suitable parameters of the proposed system are searched automatically and the algorithm is tested on the complete standard electrocardiograms database CSE, where it achieves significantly better results than other published methods.

For time-varying signals such as speech and audio, short-time analysis becomes necessary to compute specific signal attributes and to keep track of their evolution. The standard technique is the short-time Fourier transform (STFT), using which one decomposes a signal in terms of windowed Fourier bases. An advancement over STFT is the wavelet analysis in which a function is represented in terms of shifted and dilated versions of a localized function called the wavelet. A specific modeling approach particularly in the context of speech is based on short-time linear prediction or short-time Wiener filtering of noisy speech. In most nonstationary signal processing formalisms, the key idea is to analyze the properties of the signal locally, either by first truncating the signal and then performing a basis expansion (as in the case of STFT), or by choosing compactly-supported basis functions (as in the case of wavelets). We retain the same motivation as these approaches, but use polynomials to model the signal on a short-time basis (“short-time polynomial representation”). To emphasize the local nature of the modeling aspect, we refer to it as “local polynomial modeling (LPM).” We pursue two main threads of research in this thesis: (i) Short-time approaches for speech enhancement; and (ii) LPM for enhancing smooth signals, with applications to ECG, noisy nonuniformly-sampled signals, and voiced/unvoiced segmentation in noisy speech. Improved iterative Wiener filtering for speech enhancement A constrained iterative Wiener filter solution for speech enhancement was proposed by Hansen and Clements. Sreenivas and Kirnapure improved the performance of the technique by imposing codebook-based constraints in the process of parameter estimation. The key advantage is that the optimal parameter search space is confined to the codebook. The Nonstationary signal enhancement solutions assume stationary noise. However, in practical applications, noise is not stationary and hence updating the noise statistics becomes necessary. We present a new approach to perform reliable noise estimation based on spectral subtraction. We first estimate the signal spectrum and perform signal subtraction to estimate the noise power spectral density. We further smooth the estimated noise spectrum to ensure reliability. The key contributions are: (i) Adaptation of the technique for non-stationary noises; (ii) A new initialization procedure for faster convergence and higher accuracy; (iii) Experimental determination of the optimal LP-parameter space; and (iv) Objective criteria and speech recognition tests for performance comparison. Optimal local polynomial modeling and applications We next address the problem of fitting a piecewise-polynomial model to a smooth signal corrupted by additive noise. Since the signal is smooth, it can be represented using low-order polynomial functions provided that they are locally adapted to the signal. We choose the mean-square error as the criterion of optimality. Since the model is local, it preserves the temporal structure of the signal and can also handle nonstationary noise. We show that there is a trade-off between the adaptability of the model to local signal variations and robustness to noise (bias-variance trade-off), which we solve using a stochastic optimization technique known as the intersection of confidence intervals (ICI) technique. The key trade-off parameter is the duration of the window over which the optimum LPM is computed. Within the LPM framework, we address three problems: (i) Signal reconstruction from noisy uniform samples; (ii) Signal reconstruction from noisy nonuniform samples; and (iii) Classification of speech signals into voiced and unvoiced segments. The generic signal model is x(tn)=s(tn)+d(tn),0 ≤ n ≤ N - 1. In problems (i) and (iii) above, tn=nT(uniform sampling); in (ii) the samples are taken at nonuniform instants. The signal s(t)is assumed to be smooth; i.e., it should admit a local polynomial representation. The problem in (i) and (ii) is to estimate s(t)from x(tn); i.e., we are interested in optimal signal reconstruction on a continuous domain starting from uniform or nonuniform samples. We show that, in both cases, the bias and variance take the general form: The mean square error (MSE) is given by where L is the length of the window over which the polynomial fitting is performed, f is a function of s(t), which typically comprises the higher-order derivatives of s(t), the order itself dependent on the order of the polynomial, and g is a function of the noise variance. It is clear that the bias and variance have complementary characteristics with respect to L. Directly optimizing for the MSE would give a value of L, which involves the functions f and g. The function g may be estimated, but f is not known since s(t)is unknown. Hence, it is not practical to compute the minimum MSE (MMSE) solution. Therefore, we obtain an approximate result by solving the bias-variance trade-off in a probabilistic sense using the ICI technique. We also propose a new approach to optimally select the ICI technique parameters, based on a new cost function that is the sum of the probability of false alarm and the area covered over the confidence interval. In addition, we address issues related to optimal model-order selection, search space for window lengths, accuracy of noise estimation, etc. The next issue addressed is that of voiced/unvoiced segmentation of speech signal. Speech segments show different spectral and temporal characteristics based on whether the segment is voiced or unvoiced. Most speech processing techniques process the two segments differently. The challenge lies in making detection techniques offer robust performance in the presence of noise. We propose a new technique for voiced/unvoiced clas-sification by taking into account the fact that voiced segments have a certain degree of regularity, and that the unvoiced segments do not possess any smoothness. In order to capture the regularity in voiced regions, we employ the LPM. The key idea is that regions where the LPM is inaccurate are more likely to be unvoiced than voiced. Within this frame-work, we formulate a hypothesis testing problem based on the accuracy of the LPM fit and devise a test statistic for performing V/UV classification. Since the technique is based on LPM, it is capable of adapting to nonstationary noises. We present Monte Carlo results to demonstrate the accuracy of the proposed technique.

This master’s thesis deals with the automatic detection of myocardial infarction in ECG. Semester work consists of two parts. The theoretical part provides a description of the electrical conduction system of the heart, spreading of electrical activity through the heart muscle, the methods of ECG scanning and the ECG curve. There are also mentioned the causes of myocardial ischemia and various methods of its detection. Another part is devoted to high-frequency ECG, analysis of HFQRS and clinical studies which describe the use of high-frequency ECG in diagnosis of myocardial infarction. In the practical part is proposed an algorithm using low-frequency components ECG and an algorithm using high-frequency components ECG for automatic detection of myocardial infarction. The proposed algorithms are implemented in programming environment MATLAB and tested on signals from the PTB database. The final part of the master‘s thesis is devoted to the comparison of the success of myocardial infarction by means of low frequency and high frequency components of ECG and comparison of achieved results with results from clinical studies.