New/Old Net Comparison
A comparison was made between old ('Juanita') and new (medium size: serial 0491, no name as yet) EGI 128 channel nets each applied twice to a subject during a single recording session. Two one minute eyes open/closed baselines were recorded with each net in the following sequence:
New Net (first application) ec1 eo1 eo2 ec2
Old Net (first application) eo3 ec3 ec4 eo4
Old Net (second application) ec5 eo5 eo6 ec6
New Net (second application) eo7 ec7 ec8 eo8
Gains and zeros were recorded at the beginning and end of the recording session and impedences were measured for each net application. Setting an impedence limit of 50 kiliOhms bad channels were:
New Net (first application) 8 15 85 108 114
Old Net (first application) 63 75 122
Old Net (second application) 75 122
New Net (second application) no 'bad' channels
The New Net (first application) 'bad' channels had impedences of 63.6, 59.7, 68.6, 60.5 and 53.3 kiliOhms which are only slightly more than the cut off. The data from these channels looked fairly good. However, channels 75 and 122 from the old net with impedences of 308.3/339.5 and 365.9/365.0 for first/second applications respectively, showed completely bad data during both recordings. These electrodes have been added to the possibly bad list for the net.
The data were processed with raw2mat and then directly into dopower which computed the power spectra referenced to Cz. A visual inspection of the data with eegscore revealed generally very good data as shown below:
Eyes open run with new net
Eyes open run with old net
The following results were computed without any scoring so the effects of eye movements and blinks are still in the data.
First the entire power spectrum was plotted for frontal electrode sites 21 (F3) and 119 (F4) for the eyes closed and eyes open conditions. There are eight curves in each plot coded as follows in the order in which the data was collected:
New Net (first application) Solid black line first run; Dotted black second run
Old Net (first application) Solid red line first run; Dotted red second run
Old Net (second application) Solid green line first run; Dotted green second run
New Net (second application) Solid blue line first run; Dotted blue second run
So the temporal order is black, red, green, and blue with solid lines followed by dotted lines within each net application.
Eyes Closed channel 21 (F3)
Eyes open channel 21 (F3)
Eyes closed channel 119 (F4) right frontal
Eyes open channel 119 (F4) right frontal
These plots show good agreement between the new and old nets as well as a general decrease in measured power spectral density with time.
In order to look at all channels average band power density values from the .BND files were plotted versus channel. Below are plots for the Delta band (1-4Hz) and the Alpha band (8-13Hz). The color coding is again ordered black, red, green, and blue with + for new net and o for old net values, but this time there is no symbol distinction between the first and second runs within a net application.
Eyes open Delta band:
There are several interesting features evident in this graph. The broad "U" shape of the curve is due to the layout of the net with electrode 1 in the lower left front of the head with sbsequent electrodes looping around the head in a counter-clockwise direction looking down from above. Thus the lower numbered electrodes located in the front of the head are more affected by eye movements than the mid-range electrodes (from the low forties to the low one hundreds) located in the back. The net electrodes loop around the front once more, from about electrode 110 on, and are again affected by eye movement. The more localized "saw-tooth" shape (most apparent in the red and green) is due to series of 6-8 consecutive electrodes running from the lower head to near the vertex reference where a smaller potential difference is measured. However, there seems to be a pretty good overlap of measurements between the old and new nets.
Eyes closed Delta band:
Eyes open Alpha band:
Eyes closed Alpha:
