Motor Imagery Recognition of EEG Signal using Cuckoo-Search Masking Empirical Mode Decomposition

Brain Computer Interface (BCI) is a collaboration between a brain and device that enables the signals from the brain to done the external activity, i.e. Cursor, Prosthetic control or Wheel chair movement. The brain and object have the direct communication control by using BCI systems. Mostly the current research should be focused on non-invasive method. The array of neurons should be read by using the computer chips and programs then translate the signals in to action i.e., Motor Imagery (MI). The main objective is used to help the disable person without someone help. Mainly the BCI System should be very helpful for the people those who are affect from paralysis to write something and control the motorized wheel chair through thought alone. In Brain Computer Interfacing BCI) the Electroencephalogram (EEG) is a very challenging non-stationary signal. In this paper the preprocessing should be done by Least Mean Square (LMS) algorithm and Empirical mode decomposition (EMD) is a new method to extract the non-stationary signal should be apply on motor imagery recognition task. The features of EEG such as energy, fuzzy approximate entropy, Morphological features and AR coefficients are extracted using Masking empirical mode decomposition. The extracted features are selected by using the cuckoo search algorithm (CSA). In this paper the extracted features should be compare, with cuckoo search or without cuckoo search algorithm analyzed. After the feature selection features are classified by using the linear discriminant analysis (LDA) with respect to some parameters like Accuracy, Precision, Recall, Maximal (MI).