Research Article

A New Recognition Method for the Auditory Evoked Magnetic Fields

Figure 8

The neighbourhood correlation coefficients calculated from the first dataset. (a) The average neighbourhood correlation coefficient calculated by 200 single trial AEFs. The x-coordinate denotes the channel number, and the y-coordinate represents the average correlation coefficient between the selected signal and the neighbourhood signals. The red dots indicate 10 sensors located in the auditory part of the temporal lobe, which are all greater than 0.8. (b) All neighbourhood correlation coefficients calculated for 200 single trial AEFs. For each single trial AEF, 274 channels are chosen as the selected channels in turn and the correlation coefficients between them and the neighbourhood signals within their neighbourhood are calculated. The x-coordinate represents the channel number, the y-coordinate represents single trial, and the z-coordinate represents neighbourhood correlation coefficient. The red dots indicate sensors located in the auditory part of the temporal lobe, most of which have neighbourhood correlations greater than 0.8. (c) The effect comparison was performed using different values as threshold. The auditory part of the temporal lobe is demarcated by black circles. The thresholds of correlation coefficients used from left to right are 0.7, 0.8, and 0.9, respectively. The 2D images come from the first single trial AEFs from the first dataset. It can be seen that, for these single trial AEFs, 0.7 and 0.8 have similar effects, but greater than 0.9 is relatively poor.
(a)
(b)
(c)