We have used k-fold cross-validation with k ¼ 5 for testing the ability of
extracted features from IMFs. The results of the experiments are calculated with
three benchmark measures, such as accuracy, sensitivity and specificity. The measured results are tabulated in Table 1.
The results in Table 1 indicates that the features extracted from the IMFs of
empirical mode decomposition gives promising results. Among the tested features,
exponential energy feature provided better accuracy with 84%. A box-plot of
extracted exponential energy is shown in Figure 4. Katz Fractal Dimension also
provides better accuracy of 80% followed by Rényi entropy with 79%. Statistical
features and Shannon entropy gives less accuracy out of all. It is noted that Shannon
entropy giving very low sensitivity value and very high specificity, which indicates
that more number of seizure signals are miss-classified as non-seizure signals.
Feature
Accuracy (%)
Sensitivity (%)
Specificity (%)
Shannon entropy
58
26
90
Statistical features
68
54
82
Higuchi fractal dimension
71
72
70
Sample entropy
73
83
60
Approximate entropy
75
82
68
Petrosian fractal dimension
75
68
82
Rényi entropy
79
82
76
Katz fractal dimension
80
88
72
Exponential energy
84
84
84
Table 1.
Results of various features extracted from the IMFs.
Figure 4.
Boxplot of the extracted exponential energy feature of four IMFs of focal (green color box, labeled as F_IMF)
and non-focal (blue color box, labeled as NF_IMF) EEG signals.
71
Empirical Mode Decomposition of EEG Signals for the Effectual Classification of Seizures
DOI: http://dx.doi.org/10.5772/intechopen.89017