Table 8 A comparison of performances of the various methods for detection of epileptic seizures applied to the dataset from [21, 22] (reproduced from [26, 34]).
From: Combination of EEG Complexity and Spectral Analysis for Epilepsy Diagnosis and Seizure Detection
Classes | Authors (year) | Method | Dataset | Accuracy |
|---|---|---|---|---|
2 | Nigam et al. [23] (2004) | Nonlinear preprocessing filter, diagnostic artificial neural network (LAMSTAR) | A, E | 97.2 |
| Â | Srinivasan et al. [14] (2005) | Time & frequency domain features, recurrent neural network (RNN) | A, E | 99.6 |
| Â | Kannathal et al. [8] (2005) | Entropy measures, adaptive neurofuzzy inference system (ANFIS) | A, E | 92.22 |
| Â | Polat et al. [24] (2006) | Fast Fourier transform (FFT), decisiontree (DT) | A, E | 98.72 |
| Â | Subasi [25] (2007) | Discrete wavelet transform (DWT), mixture of expert model | A, E | 95 |
| Â | Srinivasan et al. [12] (2007) | Approximate entropy, artificial neural network | A, E | 100 |
| Â | Tzallas et al. [26] (2007) | Time frequency (TF) analysis, artificial neural network (ANN) | (A, B, C, D), E | 97.73 |
| Â | Ocak [27] (2008) | Approximate entropy & discrete wavelet transform (DWT), genetic algorithm(GA) | (A, B, C, D), E | 96.15 |
|  | This paper | Time frequency & approximate entropy analysis, linear or nonlinear classifiers | (A, B, C, D), E | 97.82–98.51 |
3 | Guler et al. [28] (2005) | Lyapunov exponents, recurrent neural network (RNN) | A, D, E | 96.79 |
| Â | Sadati et al. [29] (2006) | Discrete wavelet transform (DWT), adaptive neural fuzzy network (ANFN) | A, D, E | 85.9 |
| Â | Ghosh-Dastidat et al. [18] (2008) | Chaos theory and wavelet analysis, PCA, radical basis function neural network | A, D, E | 96.73 |
| Â | Mousavi et al. [30] (2008) | AR model, wavelet decomposition, MLP classifier | A, C, E | 96 |
|  | This paper | Time frequency & approximate entropy analysis, linear or nonlinear classifiers | A, D, E | 96.83–98.67 |
5 | Güler et al. [32] (2005) | Wavelet transform, adaptive neurofuzzy inference system | A, B, C, D, E | 98.68 |
|  | Güler et al. [33] (2007) | Wavelet transform, Lyapunov exponents, support vector machine | A, B, C, D, E | 99.28 |
| Â | Ãœbeyli et al. [31] (2007) | Eigenvector methods, Mixture of expert models | A, B, C, D, E | 98.60 |
| Â | Tzallas et al. [34] (2009) | Time frequency (TF) analysis, artificial neural network (ANN) | A, B, C, D, E | 89 |
| Â | This paper | Time frequency & approximate entropy analysis, RBFSVM | A, B, C, D, E | 85.9 |