This case illustrates one of the challenges of finding a physiologic response with 500 Hz tone bursts – it is hard to separate out the stimulus artifact from the physiologic response.  Let’s first look at an approximation of the stimulus – a 5 cycle Blackman tone burst.  The period of each cycle is 2 ms, so when you reverse polarity, the stimulus is “upside down”, or thought of another way, stimulus artifact would delay by 1 ms.

We typically think of the rarefaction stimulus phase as initiating the physiologic response.  The basilar membrane is pulled upward; the hair cell cilia shear towards stria vascularis: that is the excitatory phase.  At top, when the stimulus starts out as positive going (the first tiny hump goes upward: condensation), the first “large” rarefaction cycle occurs at 3 ms.  (I attempted to start the signal at -1 ms, close to the time frame of the electrical artifact.  The stimulus starts about 0.9 ms ahead of the ABR sweep so that the acoustical signal reaches the end of the insert tubing at the start of the ABR sweep: 0 ms.)

When we reverse the signal phase, the first big negative cycle is now at 4 ms.

So, if the ABR is triggered first by rarefaction and then by condensation 500 Hz stimuli, it is quite possible to see a 1 ms difference in the evoked response latencies.

Before moving on, let’s think about one other facet – the time between the major waves of the ABR: I-III and III-V are about 2 ms apart.  And the condensation peaks of a single phase 500 Hz stimulus artifact are 2 ms apart.  The response then doesn’t look too terribly different from the evoking stimulus!  See the sketch at the left; the top is of course the stimulus, the bottom trace is the ABR response.  As you’ve seen, wave V comes in later for a 500 Hz response, at some latency between about 8 and 12 ms, but the latencies are not as predictable as for click-evoked responses.

If we are looking at stimulus artifact from the two separate polarities, they will be out of phase, or thought of another way, their peaks and troughs differ by 1 ms (for a 500 Hz signal).  Let me try saying that yet another way- in the first figure above, the blue tick is 1 ms later for the signal that starts at rarefaction.

And the problem – when we change polarity of the signal, we delay the start of the ABR 1 ms because the response is initiated by the rarefaction cycle of the tone burst.

So, separating out stimulus artifact and response artifact is not so simple.  Think about what would happen if you averaged together the opposite polarity ABRs shown to the left.  Because the responses are 1 ms apart, and the peaks and valleys are 1 ms apart, averaging the two polarity responses can cancel out part of the response itself!  You can eliminate the signal artifact, but at the cost of some of the response itself!