Figure 8 illustrates the ECG artifact cancellation for the same EEG record shown in Figure 7. In its upper row (Fig. 9a), the ECG has been plotted (same x2(n) as in Fig. 8). The EEG mixed with the ECG (d2(n) in the adaptive filter cascade) is illustrated in Figure 8b. The output signal y2(n) (shown in Fig. 8b) is subtracted from the EEG contaminated with ECG artifacts, to give as result the signal e2(n), which is the EEG free of QRS complexes (Fig. 9 c).
After 50 Hz and ECG filtering, the EEG is forwarded to the third stage in order to remove EOG artifacts. Figure 9 shows the power spectrum of the EOG signal, x3(n) (Fig. 10a), the EEG mixed only with EOG, d3(n) (Fig. 10b) and the output adaptive cascade filter, e3(n) (Fig. 10c). Note that in the output signal there are no low frequencies, so indicating that the EOG was actually removed.
Table I summarizes the details for each EEG analyzed, i.e., the attenuation percentages of the components of maximum energy of line interference, ECG and EOG present in the EEG. These percentages were calculated comparing the value of the component of maximum energy of the reference signals present in the EEG records before and after filtering. In Table II the mean and the standard deviation of the previous percentages are displayed.
Figure 8 illustrates the ECG artifact cancellation for the same EEG record shown in Figure 7. In its upper row (Fig. 9a), the ECG has been plotted (same x2(n) as in Fig. 8). The EEG mixed with the ECG (d2(n) in the adaptive filter cascade) is illustrated in Figure 8b. The output signal y2(n) (shown in Fig. 8b) is subtracted from the EEG contaminated with ECG artifacts, to give as result the signal e2(n), which is the EEG free of QRS complexes (Fig. 9 c).
After 50 Hz and ECG filtering, the EEG is forwarded to the third stage in order to remove EOG artifacts. Figure 9 shows the power spectrum of the EOG signal, x3(n) (Fig. 10a), the EEG mixed only with EOG, d3(n) (Fig. 10b) and the output adaptive cascade filter, e3(n) (Fig. 10c). Note that in the output signal there are no low frequencies, so indicating that the EOG was actually removed.
Table I summarizes the details for each EEG analyzed, i.e., the attenuation percentages of the components of maximum energy of line interference, ECG and EOG present in the EEG. These percentages were calculated comparing the value of the component of maximum energy of the reference signals present in the EEG records before and after filtering. In Table II the mean and the standard deviation of the previous percentages are displayed.
การแปล กรุณารอสักครู่..
