This chapter gives an effective and efficient technique that can detect epilepsy in real time. It is low cost, low power, and real-time devices that can easily detect epilepsy. Along with EEG device, one can upgrade with GSM module to alert the doctors and parents of patients about its occurrence to prevent a sudden fall, which may cause injury and death. The accuracy of this EEG device depends on the quality of feature extraction technique and classification algorithm. In this chapter, support vector machine (SVM) is used as a classifier. Wavelet transform gives feature extraction, which helps to train data and to detect normal or seizure patients. Discrete wavelet transform (DWT) decomposes the signals into three decomposition levels. In this detection, mean, median, and non-linear parameter entropy were calculated for every sub-band as key parameters. The extracted features are then applied to SVM classifier for the classification. Better accuracy of classification is obtained using wavelet and SVM classifier.
TopIntroduction
This work consists of Electroencephalography (EEG) signals to give advanced solution to the problem related with traditional technique of visual EEG interpretation which is difficult, time consuming. The proposed technique has more significance because it’s robust nature of feature extraction. Statistical analysis of features performed in this proposed work. By using support vector machine, it get classified into normal and seizures signals (Karumari et al., 2016). This technique has one importance of robust and efficient classification of epileptic signals.
Epilepsy is a dangerous brain disorder which relate to the sudden seizures which causes immediate change of the brain electrical activity which needs proper diagnose to minimize its occurrence before the condition is converting into worst situation (Saminu, 2019), (Jaiswal & Banka, 2017). One can perform training of stored data to check detection first and then experimentation performed on real time database. One of the most important and challenging task of epilepsy detection is selection of proper classifier and nature of extracted features (Wang et al., 2016). As part of signal processing feature extraction can be performed by different transforms like wavelet, Fourier transform, canny edge algorithm and many more. Also for classification one can use SVM, ANN, K-NN, K means clustering algorithms which gives different accuracies (Willems et al., 2019). From the signal processing point of view, several techniques have been in existence for epileptic detection and classification. These techniques are developed based on different types of domains like time, frequency, time frequency (Adeli et al., 2017). It was proved in literature that features extraction from frequency sub-bands gives higher important information more than those from entire EEG signals. Some of the works completed has either low accuracy, used a single domain or linear features while some employed low efficient classifiers (Mathieson et al., 2016), (Fasil & Rajesh, 2019). Classification efficiency can be improved using SVM classifier and feature extraction completed using SVM classifier.
Epilepsy is a neurological disorder that affects approximately fifty million people according to the World Health Organization (WHO). While EEG plays important roles in monitoring the brain activity of patients with epilepsy and diagnosing epilepsy, an expert is needed to analyze all EEG recordings to detect epileptic activity. This method is obviously time-consuming and tedious, and a timely and accurate diagnosis of epilepsy is essential to initiate antiepileptic medicine therapy and subsequently reduce the risk of future seizures and seizure-related complications.
In this study, a convolutional neural network (CNN) based on raw EEG signals instead of manual feature extraction was used to distinguish ictal, preictal, and interictal segments for epileptic seizure detection. This work compares the performances of time and frequency domain signals in the detection of epileptic signals based on the intracranial Freiburg and scalp CHB-MIT databases to explore the potential of these parameters. Three types of experiments involving two binary classification problems (interictal vs. preictal and interictal vs. ictal) and one three-class problem (interictal vs. preictal vs. ictal) were conducted to explore the feasibility of this method. Using frequency domain signals in the Freiburg database, average accuracies of 96.7, 95.4, and 92.3% were obtained for the three experiments, while the average accuracies for detection in the CHB-MIT database were 95.6, 97.5, and 93% in the three experiments. Using time domain signals in the Freiburg database, the average accuracies were 91.1, 83.8, and 85.1% in the three experiments, while the signal detection accuracies in the CHB-MIT database were only 59.5, 62.3, and 47.9% in the three experiments.
Based on these results, the three cases are effectively detected using frequency domain signals. However, the effective identification of the three cases using time domain signals as input samples is achieved for only some patients. Overall, the classification accuracies of frequency domain signals are significantly increased compared to time domain signals. In addition, frequency domain signals have greater potential than time domain signals for CNN applications.