Sharif Digital Repository

Brain computer interfaces (BCI) are systems designed to understand the brain function from activation patterns and dynamics recorded from the brain activity and use this knowledge to give disabled people the ability to communicate with their surroundings. Features are extracted from recorded signals from the brain while occupied in a mental task and classified into categories related to the task given. These classifiers are then used for the estimation of user anticipation. Usually, the tasks defined are meant to evoke or induce a potential in the pattern of the brain. Awareness of error responses is one of the cognitive functions of the brain which occurs when a response is in conflict with... 

Even in the simplest of activities in the brain such as resting condition, there are connections in between different regions of the brain so that the whole system functions consistently in harmony. Studies related to brain connectivity provides an opportunity to better understand how the brain works. To assess these connectivities an estimation is usually conducted based on brain signals. Among different estimation methods, quantities of information theory are in general more practical due to avoiding any assumptions toward the system model and the ability to recognize linear and non-linear connectivity. One of the main quantities related to the information theory is in fact, entropy.... 

The human brain can be described as a dynamic system with multiple subsystems interacting with one another and multi-channel observations of these subsystems are available. The modeling of a system from its observations allows us to gain insight into how its various components interact with one another and also provides intuition about the desired system. Hidden Markov Model (HMM) is a probabilistic model with hidden states that is suitable for modeling these types of systems. Multi-channel observations are available from several subsystems interacting with each other within a general system. In this case, it may be necessary to develop more comprehensive models incorporating multi-channel... 

Epilepsy is one of the most common neurological disorders, second only to stroke, with a prevalence of 0.6–0.8% of the world’s population. Epilepsy is not cured, Two-thirds of the patients achieve sufficient seizure control from anticonvulsive medication, and another 8–10% could benefit from respective surgery. For the remaining 25% of patients, no sufficient treatment is currently available. In the recent decent, many studies in this field attempt to predict the onset time of a impending seizure by monitoring the biomedical signals, Electroencephalogram (EEG) Signal processing is one of these biomedical Signals that many In this work, First we talk about the basic concepts of Epilepsy ,... 

The aim of this project is designing and implementation of a BCI for interacting with virtual environments. BCI is an interface between brain and computer that connect those by translating brain signals into understandable instruction for computer. There are different ways to generate brain states for classification in BCI technology. One of those, is based on motor imagery (MI). In this project, we used motor imagery of left hand, right hand and foot. subject training in this kind of BCI (MI-based BCI) is very vital. in order to speed up subject training period and consequently enhance the separability of brain states, computer graphics and virtual environments are used to excite the... 

Brain is the most complicated organ of body which controls the activity of all other organs. Understanding its function and its language could give us a direct communication pathway for connecting injured motor organ and it could be useful for functional repairing. Neurons are atoms of a vast network that generate the brain signals. Processing these signals would help to translate brain’s language and has three main stages: spike detection from signal, spike sorting, and intention extraction from encoded signal. In this research, we use a dataset of rat’s extracellular recordings during a time interval in which a rat pressed the liver several times to receive water as an award. Scince spikes... 

About 42 percent of annual mortality in all around the world is originated from cardiovascular arrhythmias and diseases. One of these arrhythmias is atrial fibrillation whose onset and persistence can produce clot and consequently cause stroke. The basis of our research are upon this idea that dangerous heart arrhythmias do not happen abruptly and there always are some background signs before occurrence of them. In our approach to predict the onset of atrial fibrillation, by analyzing ECG signal in order to extract distinguishing features, we want to classify signals which will terminate Paroxysmal Atrial Fibrillation (PAF) from signals which won’t end with PAF. In this thesis, we propose... 

Brain is the most complicated organ of body which controls the activity of all other organs. Understanding its function and its language could give us a direct communication pathway for connecting injured motor organ and it could be useful for functional repairing. Neurons are atoms of a vast network that generate the brain signals. Processing these signals would help to translate brain’s language and has three main stages: spike detection from signal, spike sorting, and intention extraction from encoded signal.
In this research, we use a dataset of rat’s extracellular recordings during a time interval in which a rat pressed the liver several times to receive water as an award. Since... 

Apnea-bradycardia is a medical term for prolonged respiratory pause accompanied with a heart rate reduction which is a common event among preterm infants. Repetition of apnea-bradycardia episodescompromises oxygenation and tissue perfusion and may lead to neurological impairment or even short-term morbi-mortality. Main solution to this breathing-related disorder is continues monitoring of infants in neonatal intensive care units in order to detect apnea-bradycardia event, generate an alarm and warn available nurse or physician to initiate quick nursing actions. Various studies have been done in this area and different methods are proposed which mainly focus on cardiac signal processing. This... 

Noise contamination is inevitable in biomedical recordings. In some cases biomedical recordings are highly contaminated with artifacts which make the effective recovering process hard to achieve. Many different methods have been proposed for artifact removal from biomedical signals but introducing an effective method which can present valuable data for medical analysis, is still an ongoing process.
This dissertation focuses on inter-ictal EEG denoising approaches including ICA-based and EMD-based methods and different combination of these methods. These methods are tested on simulated epileptic recordings which are contaminated with real muscle artifact and EEG signal. The denoised... 

In this work, we evaluate differernt tensor decomposition methods in application of fECG extraction from abdominal ECG recordings. After selecting proper tensor decomposition tool (Tucker decomposition) we propose a linear source separation algorithm based on a measure of quasi-periodicity. The quasi-periodicity is attained through the use of a constraint on a matrix factorization problem. In practice, we form a three dimensional ”tensor” by stacking the observation matrix and rough estimates obtained by both linear and non-linear subspace reconstruction methods. The method is applied to a database of electrocardiography (ECG) recordings, where rough subspace estimates of maternal and fetal... 

Cardiovascular diseases are one of the major causes of mortality in humans. One way to diagnose heart diseases and abnormalities is processing of cardiac signals such as ECG.In many of these processes, inter-beat and intra-beat features of ECG signal must be extracted. These features include peak, onset and offset of ECG waves,meaningful intervals and segments that can be defined for ECG signal. ECG fiducial point (FP) extraction refers to identifying the location of the peak as well as the onset and offset of the P-wave,QRS complex and T-wave which convey clinically useful information. However, the precise segmentation of each ECG beat is a difficult task, even for experienced... 

Respiration signal is one of the biological information required to monitor patient respiratory activities. Noninvasive respiratory monitoring is an extensive field of research, which has seen widespread interest for several years. It is well known that the respiratory activity influences electrocardiographic measurements (ECG) in various ways. Therefore, different signal processing techniques have been developed for extracting this respiratory information from the ECG, namely ECG derived respiratory (EDR). Potential advantages of such techniques are their low cost, high convenience and the ability to simultaneously monitor cardiac and respiratory activity. One of the aims of this thesis is... 

Transfer Learning is one of the most important fields in the Machine Learning area. Respect to the advances that we have seen in the Computer Science, especially in the Machine Learning area, we need a tool that can transfer learnings from different domains to each other. As data distribution varies, many statistical models require restructuring using new training data. In many applications, re-assembling training data and re-structuring models is inefficient and costly, so reducing the need for this practice seems appropriate. In these cases, knowledge transfer or learning transfer between domains may be desirable. For example, in the area of the B rain-Computer Interface, when it... 

Objective: Joint analysis of multi-subject brain imaging datasets has wide applications in biomedical engineering.In these datasets, some sources belong to all subjects (joint), a subset of subjects (partially-joint), or a single subject (individual). In this thesis, this source model is referred to as joint/partiallyjoint/individual multiple datasets unidimensional (JpJI-MDU), and accordingly, a source extraction method is developed.Method: We present a deflation-based algorithm utilizing higher order cumulants to analyze the JpJI-MDU source model. The algorithm maximizes a cost function which leads to an eigenvalue problem solved with thin-SVD (singular value decomposition) factorization.... 

The human brain is one of the most complex and least understood systems in nature. In recent decades, numerous studies have been conducted to identify the behavior of this system. One of the areas of brain research is the investigation of the connections between different regions of the brain during a presumed process or in a resting state. Among various types of brain connections, effective connectivity provides researchers with higher-level information on brain behavior compared to other connections, but also entails greater computational complexity. In recent years, researchers have aimed to provide an estimator with the maximum desirable capabilities, and with the advent of (deep) neural... 

Stress is an integral aspect of human experiences, manifesting in various situations such as tension, anxiety, and specific diseases. Researchers globally are exploring innovative approaches to control stress, aiming to quantify and detect it. The quantification of stress levels using brain signals is a prominent focus in the field of medical engineering.In most studies, the classification objective revolves around two classes (stress-relaxation) or three classes (stress levels). Typically, these studies employ databases inducing stress and recording brain signals uniformly. While various stress induction methods have been used, the resulting brain connections are less discussed.This project... 

There are several methods for EEG state detection, and there are still many challenges. Switching Kalman Filter (SKF) is a suitable approach for state detection, which has been used in various applications such as QRS detection in ECG signal, apnea detection using ECG signal, and also hand path detection using EEG signal. Our goal here is to use Switching Kalman Filter (SKF) in order to detect changes in EEG signal, and in particular in sleep. In other words, we want to detect Sleep Stages. Here, detecting Sleep Stages will help doctors diagnose and treat diseases. There is a Kalman Model for each Stages of Sleep in SKF, that I model it with a AR model. In addition, SKF switch is a state... 

Interictal EEG signals are very critical in diagnosis of epilepsy. Analysis of interictal EEG signals is very challenging due to contamination by various undesired signals like background EEG, muscular activity, noise, etc. Thus denoising of interictal signals has been an active research field in recent years. Primary purpose of this thesis is to denoise interictal EEG signals by using different combinations of ICA-based and wavelet denoising approaches. Then a new direction is pursued by using Morphological Component Analysis (MCA) which is a method for solving source separation problems based on morphological diversity of sources. Afterward MCA is modified by considering more prior... 

BCI is an interface between brain and machine, particularly computer which translates brain signals into understandable instructions for machine. BCI records signals and determines what the subject is doing or thinking. BCI in the point of view of pattern recognition is a classification problem. For this aim, different tasks are referred to different classes. The more number of classes, the higher complexity we encounter in classification so surveying of different kinds of features, feature selection and reduction methods have highly importance. In this project we want to design a 4-class classification that each class is referred to a direction of wrist movement. During the time that the...