Browsing by Author "Rajaguru, H"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    item: Conference-Full-text
    Factor analysis, hessian local linear embedding and isomap for epilepsy classification from EEG
    (Institute of Electrical and Electronics Engineers, Inc., 2016-12) Prabhakar, SK; Rajaguru, H; Rajapakse, A; Prasad, WD
    In this present generation, more than 1% of the whole world’s population is affected by this seizure disorder. Due to genetic predispositions like tumours, strokes and drug misuse, epilepsy is caused. Epilepsy is a common brain disorder in which cluster of neuronal signals function abnormally. For preventing seizures, medications are available but only some patients can respond well to the medication. Other remedial measures such as neurostimulation, surgery, maintaining a healthy diet are not always successful in treating the patient. Because of epilepsy, the patients have to live with persistent anxiety throughout their lives and also leading a normal life and performing day to day and social tasks becomes more difficult for them. For the analysis and diagnosis of epilepsy, the detection and classification of seizures forms the most important step. The vital information regarding the dynamics of the brain can be easily measured by Electroencephalogram (EEG). Since the recordings of the EEG data are pretty long, the obtained data is too huge to process and so, in this study, the dimension of the EEG data is reduced by Dimensionality Reduction (DR) techniques such as Factor Analysis (FA), Hessian Local Linear Embedding (HLLE) and Isomap (IM). The dimensionally reduced values are then classified with the help of Genetic Algorithm (GA) and Generalized Regression Neural Network (GRNN). The performance metrics are analyzed with parameters such as performance index, sensitivity, specificity, time delay, quality value and accuracy. The results show that when FA, HLLE and Isomap are classified with GRNN, then a perfect classification of 100% and an accuracy of 100% are obtained. If the dimensionality reduction techniques classified with GA are compared, then a high accuracy of 95.24% is obtained when it is classified with HLLE-GA combination.
  • Thumbnail Image
    item: Conference-Abstract
    A simplified epilepsy classification technique utilizing svd
    (Engineering Research Unit, Faculty of Engiennring, University of Moratuwa, 2016-04) Prabhakar, SK; Rajaguru, H; Jayasekara, AGBP; Amarasinghe, YWR
    EEG signals represent both the brain function and also the status of the whole body, i.e. a simple action as blinking the eyes introduces oscillation in the EEG records. The EEG is a direct way to measure neural activities and it is important in the area of biomedical research to understand and develop new processing techniques. EEG signal pre-processing and postprocessing methods include EEG signal modeling, segmentation, filtering and de-noising, and EEG processing methods which consist of two tasks, namely, feature extraction/dimensionality reduction and classification. In this paper, the performance analysis of Independent Component Analysis (ICA) is considered as a dimensionality reduction technique followed by Singular Value Decomposition (SVD) as a Post Classifier for the Classification of Epilepsy Risk Levels from EEG Signals. The analysis is done in terms of bench mark parameters such as Performance Index (PI), Quality Values (QV), Sensitivity, Specificity and Time Delay.