Effects of Electrode Placement
From an electrical engineers perspective, EEG signals originate from the summation of a large number of events where small voltage pulses are generated by electrochemical activity. Each pulse can be seen as an electrical dipole having a vector direction. The electrical energy of the pulse travels through the conductive tissue and fluids of the brain, through the skull, scalp, and to our electrodes. The head can be modeled as a "volume conductor", with essentially equal conductance throughout. Electrical signals travel at the speed of light in a volume conductor.
There are other neurological pathways that transmit information in an electrochemical chain which are beyond the scope of this paper. From an EEG perspective, we can only measure that activity which results in significant electrical energy being released into the volume conductor.
With modern instrumentation, the ground electrode plays no significant part in the measurement. It is only necessary to provide "electronic housekeeping" for the amplifiers. Therefore it can be placed anywhere on the body. We use a conductive rubber wrist strap electrode for convenience.
The following test was performed to measure the effects of electrode placement on signals originating in various parts of the brain.
A fish bowel approximately the size of a human skull was outfitted with silver chloride electrodes epoxied to the sides of the bowel. Placements were at simulated left ear, right ear, and Cz. Additional electrodes simulated a narrow bipolar placement such as Cz-C4, and a wider placement such as C3-C4. The bowel was filled with a saline solution approximating that of biological fluids.
An electronic signal source of approximately 10 Hz was connected to a "dip stick" consisting of a plastic rod with two silver chloride buttons on either side. This allowed simulating a dipole type signal with a know location and vector direction in the fluid.
Each set of measurements in the table below was performed in the same way. A set of electrode were chosen to simulate the required placement and connected to the F1000 preamplifier. The F1000 EEG filter was set to a bandwidth of 8-12 Hz. The dip stick was then moved to various positions in the bowl and the resultant readings recorded.
The results are shown graphically in the table with comments as appropriate.
direction indicates dipole direction for maximum pickup.
Arrow size indicates sensitivity to signals in that area.
This type of placement emphasizes the area of the head immediately under the electrodes. It is most sensitive to dipoles oriented inline between the electrodes.
Signal amplitude will generally be lower than other placements due to the small area of the head being measured.
|Bipolar, Wide Spaced
The two main differences in this placement is the deeper penetration into the head, and the tendency to emphasis the areas immediately under the electrodes.
Ear to Cz
Notice that this placement is really the same as bipolar wide spaced, except that the ear is one of the electrodes. While not shown here, note that moving the Cz electrode a short distance would probably not produce a significantly different result.
Ears to Cz
Tends to spread pickup to the whole head. Again we get some emphasis of the areas just under electrodes. This small difference probably explains why we are able to do scanning with this configuration. With statistical analysis the small differences can be made significant. Of course, anything else that produces small differences would also affect the measurements.
Last Changed 05/19/04