Corrigendum to 'Blink-related delta oscillations in the resting-state EEG: A wavelet analysis' (Neurosci. Lett. 449 (2009) 57-60)

摘要

Subsequent recordings collected during a follow-up of our studies on blink-related EEG have not produced results overlapping those reported in the previous paper [3]. In particular, the blink-related delta oscillation, that was a biphasic negative-positive waveform in the above-mentioned paper, has turned out to be reversed in polarity (i.e., positive-negative) in the recent recordings. The only difference in the EEG data processing with respect to the previous recordings was in the export file format: EDF (recent) vs ASCII (previous). We believe that the polarity originating from the EDF files (i.e., positive-negative) should be considered correct, because EEG data exported in the form of EDF files, according to the specifications of the format itself, faithfully represent original raw data [6,7]. On the contrary, data exported in the form of file text (with an ASCII code) only allow reproducing signals visualized on the recording device screen. But, in doing so, the waveform is reconstructed referring to an internal code, according to which negative values are assigned to the points situated in the inferior part of the picture (with respect to an equatorial line of symmetry). This poses a serious problem when the reference coordinates of the recording device, concerning vice versa the polarity of the deflections with respect to the isoelectric line, are such that positive deflections are downward. In this case, in fact, the procedure of exporting data as ASCII files, by assigning negative values to the same deflections (because they are really downward from a spatial point of view), gives rise to an error of inverted polarity. In other words, such a procedure preserves the graphic representation of the data without considering the true polarity.