I.J. Modern Education and Computer Science, 2014, 7, 31-39 Published Online July 2014 in MECS (http://www.mecs-press.org/) DOI: 10.5815/ijmecs.2014.07.05 Copyright © 2014 MECS I.J. Modern Education and Computer Science, 2014, 7, 31-39 Automatic Removal of Artifacts from EEG Signal based on Spatially Constrained ICA using Daubechies Wavelet Vandana Roy Department of Electronics & Communication, GGITS, Jabalpur, M.P., 482005, INDIA vandana.roy20@gmail.com Dr.Shailja Shukla Professor & Head of Department of Computer Science Engineering, JEC, Jabalpur, MP, 482002, INDIA shailja270@gmail.com Abstract—This paper presents a boon and amend technique for eradicating the artifacts from the Electroencephalogram (EEG) signals. The abolition of artifacts from scalp EEGs is of considerable implication for both the computerized and visual investigation of fundamental brainwave activities. These noise sources increase the difficulty in analyzing the EEG and procurement clinical information related to pathology. Hence it is critical to design a procedure for diminution of such artifacts in EEG archives. This paper uses a blind extraction algorithm, appropriate for the generality of complex-valued sources and both complex noncircular and circular, is introduced. This is achieved based on higher order statistics of dormant sources, and using the deflation approach Spatially-Constrained Independent Component Analysis (SCICA) to separate the Independent Components (ICs) from the initial EEG signal. As the next phase, level-4 daubechies wavelet db- 4 is applied to extract the brain activity from purged artifacts, and lastly the artifacts are projected back and detracted from EEG signals to get clean EEG data. Here, thresholding plays an imperative role in delineating the artifacts and hence an improved thresholding technique called Otsu’s thresholding is applied. Experimental consequences show that the proposed technique results in better removal of artifacts. Index Terms—Artifacts removal, Biomedical Signal Filtering, Electroencephalogram (EEG), source separation, Spatially-Constrained Independent Component Analysis (SCICA), thresholding, daubechies wavelet. I. INTRODUCTION Human brain possesses rich spatiotemporal subtleties Because of its complicated uncertain cautious nature. Electroencephalography (EEG) provides a direct determination of cortical behavior with millisecond temporal steadfastness when compared to supplementary techniques. Electroencephalogram (EEG) is multivariate time series data measured using multiple sensors positioned on scalp that imitates electrical potential produced by behaviors of brain and is a record of the electrical potentials created by the cerebral cortex nerve cells. There are two categories of EEG, which is based on location of the signal obtained in the head: scalp or intracranial. Scalp EEG as being the main focus of the research, uses small metal discs, also known as electrodes, which are kept on the scalp with good electrical and mechanical touch. Intracranial EEG is obtained by special electrodes placed in the brain during a surgery. The electrodes should be of minimum impedance, in order to record the exact voltage of the brain neuron. The variations among the voltage difference among electrodes are sensed and amplified before being transmitted to a computer program. Electrical impulses generated by nerve firings in the brain diffuse through the head and can be measured by electrodes placed on the scalp, & is known as electroencephalogram (EEG). The artifacts, such as eye blinks etc., in EEG recordings obscures the underlying processes and makes analysis difficult. Large amounts of data must often be discarded because of contamination by artifacts. To overcome this difficulty, signal separation techniques are used to separate artifacts from the EEG data of interest. The noise, or artifacts, sources include: line noise from the power grid, eye movements, eye blinks, heartbeat, breathing, and other muscle activity. Some artifacts, such as eye blinks, produce voltage changes of much higher amplitude than the endogenous brain activity. In this situation the data must be discarded unless the artifacts can be removed from the data. EEG data may be contaminated at many points during the recording and transmission process. Most of the artifacts are biologically generated by sources external to the brain. Improving technology can decrease externally generated artifacts, such as line noise, but biological artifacts signals must be removed after the recoding process. Figure 1 shows waveforms of some common EEG artifacts.