There are two type of damage that can result from head injuries: damage to gray matter (the neuron cell bodies) or matter to white matter (the neuron connective filaments).

Gray matter damage

Gray matter damage results in loss of neurons, no matter what speed they were firing, in an area. Areas of very low amplitudes overall relative to surrounding and homologous areas relate to gray-matter damage, literally a killing of the neurons, so there are many fewer producing the electrical pulses we read as EEG.  Comparing one site with its mirror site (say, C3 compared with C4) you might see a significant difference in the amplitudes at most or all frequencies, with the low side being the damaged side. It is my understanding that over time many brains can repopulate these areas, so gray-matter injuries can resolve with time.  Low amplitudes on one side suggest a gray-matter head injury at those sites.

Alpha spikes secondary to head injuries (gray matter) are common.  As damaged neurons are replaced they often lock into alpha without the connections the old ones had. If EO alpha is high–higher than theta and delta–then there is a problem with alpha blocking–no matter how much it has come down.  I would look for EO Alpha Theta ratios in back at or below 1.0.  As I’m sure you know, a blow to the back of the head can often result in a contra-coup injury, where the brain is slammed forward into the skull on the opposite side–the prefrontal in this case.

When alpha shows a spike at a site relative to the rest of the head, one of the first things to rule out is a closed head injury.  Gray matter damage (neurons, not axons) results first in reduced amplitudes in all frequencies (fewer neurons to fire), but as the neurons are replaced the amplitude comes back—though the connections don’t—and it’s not uncommon to see a spike of alpha.

White matter damage

White matter damage doesn’t kill the neurons; it just cuts them off from the rest of the world–like a whole class of high school students all losing their cellphones. With no incoming or outgoing communications, the cell tends to drop to the lowest frequency of the brain–the incommunicado frequency we see in deepest sleep–Delta (and often very low theta).  Areas of very high delta relative to surrounding and homologous areas relate to white-matter damage, a shearing of the connections, which leaves neurons as little islands of activation with no inputs or outputs. The delta is projected from the brainstem when nothing else is happening.

The problem with axonal shearing is not loss of the myelin sheath.  It is the tearing of the axon itself. The sheath makes it possible for axons to carry faster pulse rates, assuming an intact axon, the “cable” coming out of the neuron to connect it with others downline. But if the axon is torn apart, then the neuron loses its ability to communicate with neurons with which it used to connect.  I can’t say whether or not the brain can replace myelin, but to the best of my knowledge–short stem cell or some other new and as-yet unproven approach–nothing creates new axons.  Learning expands the trees of dendrites which receive signals from other axons and expands the links off the axon to allow it to connect with other downline cells.  But the axon, once it is gone, is gone to the best of my knowledge.

Of course, the amplitudes of these frequencies would be HIGHER on the side where the damage occurred.  And white matter damage does not heal.  I can’t find the source now (I’m pretty sure it was Jay Gunkelman), but I learned several years ago that these neurons over time can tend to shift up into the low alpha frequency, resulting in high amplitudes of 8-10 Hz instead of Delta.

Training for brain injury

One of the most basic rules of training injured brains is that you do NOT train right over the injury.  Work around rather than directly over the affected area.  I have never found a huge benefit in localizing the level of dysfunction to a particular sub cortical structure, lobe or 10/20 site. In most cases it’s really NOT that localized.  So, training the whole left hemisphere or the left-frontal quadrant tends to require the brain to work the area that isn’t working so hot AND the surrounding areas that can pick up some of the load.

I would think that wherever you have significant slow activity, which should show up on the assessment on the heads page, you would be very safe just to work with reducing slow activity. If you have fairly broad areas of slowing, then use a 2 summed channel protocol just reducing the most problematic band. 2-5 Hz is a pretty safe frequency, which is not likely to further impact memory.

Damage to the left frontal cortex (around F3), resulting in excessive slow activity there relative to right, often results in depression; conversely, damage to the right frontal cortex (around F4), resulting in excessive slow activity there relative to left, often results in mania.  It would seem that a true bipolar might shift back and forth between overactive left and then right frontal areas.

With someone who has had not one but two aneurisms, I’d probably stay away from pushing faster speeds, but certainly SMR has been used very effectively with seizures, especially between C3 and C4 or at Cz in one channel.

Excess coherence often follows a head injury.