3D Phase Space Reconstruction Research Articles

Higher-Order Phase-Space Reconstruction for Detection of Epileptic Electroencephalogram

In this paper, the authors propose a new technique for the classification of seizures, non-seizures, and seizure-free EEG signals based on non-linear trajectories of EEG signals. The EEG signals are decomposed using the EMD technique to obtain intrinsic mode functions (IMFs). The phase space of these IMFs is then reconstructed using a novel technique of higher-order dimensions (3D, 4D, 5D, 6D, 7D, 8D, 9D, and 10D). The existing techniques of seizure detection have deployed 2D & 3D phase–space reconstruction only. The Euclidean distance of all higher-order PSR is used as a feature to classify seizures, non-seizures, and seizure-free EEG signals. The performance of the proposed method is analyzed on the Bonn University database in which 7D reconstructed phase space classification accuracy of 99.9% has been achieved both using Random Forest classifier and J48 decision tree.

Discrimination of multi-class EEG signal in phase space of variability for epileptic seizure detection using error correcting output code (ECOC)

A new approach for detection of epileptic seizure based on multiclass decision tree classier combining with error correcting output codes (ECOC) using ensemble learning is presented in this paper. The dynamical properties in electroencephalogram (EEG) recorded data of electrical character of the brain during seizure is different from a normal behavior of the brain. Based on the electrical characteristics of the brain, the classification of epileptic seizure in EEG signal is performed in two steps: extraction of feature by first-order difference (FOD) whose variability has been used in 95% confidence ellipse area computation in 2D phase space reconstruction (PSR) and Interquartile range (IQR) of Euclidean distance of FOD plot in 3D PSR; and for classification, bagging algorithm of ensemble learning with decision tree as a base classifier and ECOC is used. This shows a great discriminating ability among datasets belonging to five groups (denoted A–E) at different state of the brain. The proposed method shows 100% classification accuracy.