How to develop a testing strategy for aided cortical testing

This video will explore how to perform aided cortical testing, using what we know about each patient to determine the best testing strategy for their unique needs. For aided cortical testing, we cannot adopt a "one size fits all" approach when it comes to deciding which stimuli and presentation levels should be tested. Each individual patient will present with different audiograms, and therefore different clinical questions. The aided cortical test can be a valuable tool to help with counselling and so it is important to bear in mind how the test can best be used to serve the needs and questions of parents / carers.

You can read the full transcript below.

Starting stimulus and level

When performing the aided cortical test, it is important to adopt a testing strategy that is appropriate and suitable for each individual patient.

The first consideration should be around which starting stimulus and which intensity level should be used for the test.

The results of this first initial test can then be used to inform subsequent testing as the assessment progresses.

ManU-IRU stimuli

Let us refer back to the ManU-IRU stimuli set.

Within this set are three stimuli, a low-frequency, a mid-frequency, and a high-frequency stimulus.

We know that we have excellent frequency specificity with these stimuli, so we can be confident that any cortical responses recorded are being generated by neural firing triggered by the region of the basilar membrane that corresponds to the frequency region associated with each stimulus.

We also know that the ManU-IRU stimuli have been calibrated to align with the ISTS.

The ISTS is representative of natural running speech, so we know that if we present any of the ManU-IRU stimuli at a level of 65 dB Speech Reference Level, then we are presenting these sounds as they would be spoken, and of course as a result heard, within a typical conversational speech level.

What is important to note here is that this does not mean each of these stimuli is presented at 65 dB SPL.

As we can see on the graph, the high frequency stimulus presented at a level of 65 dB Speech Reference Level actually peaks at just under 45 dB SPL.

On first glance, this may seem as though this stimulus is being presented at an excessively quiet level, but this is the key to understanding the ManU-IRU stimuli.

When a person is speaking at a typical conversational speech level of 65 dB, those high-frequency sounds within that speech are actually a lot quieter than 65 dB, and a lot quieter than the mid- or low-frequency sounds.

Here we can see the effective offsets between the Speech Reference Level and the actual sound pressure levels.

This is the difference between the ISTS level and the level of each individual ManU-IRU stimulus.

We can see the biggest difference is in the high frequency, which makes sense because these are the quietest sounds that will be found within natural speech and within the ISTS.

For a 75 dB Speech Reference Level, in other words, a presentation of the ManU-IRU stimuli at 75 in the software, which corresponds to loud natural speech.

The actual dB SPL levels are 72.3, 60.5, and 54.4 for each of the stimulus tokens.

For a 65 dB Speech Reference Level, in other words a presentation of the ManU-IRU Stimuli at 65 in the software, which corresponds to conversational natural speech, the actual dB SPL levels are 62.3, 50.5, and 44.4 for each of the stimulus tokens.

For a 55 dB Speech Reference Level, in other words a presentation of the ManU-IRU stimuli at 55 in the software, which corresponds to quiet natural speech, the actual dB SPL levels are 52.3, 40.5, and 34.4 for each of the stimulus tokens.

Mild losses

One important patient cohort to give special consideration to is those with mild losses.

In these cases, the main question is, might they be able to hear the sounds unaided?

Because if so, they will definitely be able to hear the stimuli aided, and so there's minimal utility to performing the aided cortical test, as the results won't demonstrate any additional aided benefit over their unaided test results.

For these mild losses, we can establish if they are likely to be able to hear the sounds in the unaided condition by referring to visible speech mapping, which can demonstrate a patient's audibility in both the aided and unaided condition.

Here, we can see the visible speech mapping screen for the unaided scenario for a mild loss.

Overlaid on this screen is the 55 dB ISTS with the 30th and 99th percentile of the variation of the speech sounds shown.

To determine how much of the signal at 55 dB would be audible for this patient without their hearing aids, we can look at how much of the signal is above their threshold.

Here is the threshold highlighted by the green line.

In visible speech mapping, the threshold is shown converted into dB SPL.

It is possible to see that some low and mid frequency content is above the threshold and therefore audible at 55 dB, but the high frequency content isn't.

It's important to highlight that although some of the low frequency content appears to be audible, it is a very small amount of the full range of speech being spoken within that ISTS presentation.

As a result, it may be valuable in this case to perform the cortical test, as very little of the test signal is going to be audible unaided, and so aided benefit may be demonstrated by the results of the test in the aided condition.

If the patient's hearing loss is incredibly mild, then the 55 dB stimuli may be audible unaided.

It is important to question whether the aided results are going to provide useful clinical information if benefit cannot be demonstrated.

If the hearing loss is closer to the moderate end of the mild range, then the 55 dB stimuli may be inaudible when unaided.

It is recommended to refer to the visible speech mapping curves to establish how much of the stimuli are below threshold.

If the majority of the stimuli appear to be below threshold, then there is potential value in performing the cortical test as aided benefit may be demonstrated.

However, there is an argument that it can be beneficial to perform the aided cortical test for all patients with mild losses to ensure aided access to speech sounds is being provided via their device and to demonstrate that the occlusive elements of the ear mold are not obscuring sounds from being detected and to ensure that the low gain provided by the hearing aids is sufficient.

When considering whether to test mild losses and whether to use the 55 dB Speech Reference Level intensity, it is important to bear in mind the likely noise floor within the testing room.

If it is not a soundproof booth, then ambient noise measurements are recommended to establish whether the noise floor is greater than the presentation level, and this may be further exacerbated by a noisy or unsettled patient.

Expected aided audibility

So how do we use what we know about the ManU-IRU stimuli to determine which stimulus and level to start testing with?

To answer this question we need to consider what the expected aided audibility is likely to be for each individual patient for the different stimuli.

To establish what the expected aided audibility is we can refer to the real ear measurement traces from the hearing aid verification process.

And here is another reason why the ManU-IRU stimuli are particularly effective.

Hearing aid verification uses the ISTS to inform the target gain for the different input levels for the device.

When we use the aided cortical test to validate the hearing aid fitting, what we are essentially asking is, has the hearing aid fitting provided sufficient gain to ensure the ISTS and all the sounds within it are audible?

Where the ManU-IRU stimuli match the levels within the ISTS for the particular speech tokens, we are able to check this accurately.

When reviewing the REM traces, if the match to target, the hearing aid output, in the frequency region that corresponds to the individual ManU-IRU stimuli.

If that match to target is more than 10 dB above the patient's threshold, then we can predict that this speech sound is likely to be audible to the patient and is therefore likely to elicit an aided cortical response.

From here, we can determine which stimuli are likely to be audible at which levels and then start our testing based on that information.

If the hearing aid output is less than 10 dB above threshold, then it is less likely that an aided cortical response will be elicited.

Using the 65 match to target is a good starting point.

If audibility is likely based on the hearing aid output being greater than 10 dB above threshold, then testing at a 65 dB speech reference level is an advisable initial testing level.

If the hearing aid output at 65 is less than 10 dB above threshold then a cortical response is unlikely to be elicited.

The next step is to refer to the 75 REM traces and if the match to target at this level is greater than 10 dB above threshold then testing can begin at this presentation level.

Remember this needs to be checked for the individual stimuli regions across the REM traces.

If we take this example, firstly we can see the patient's threshold denoted by the edge of the lower shaded area.

Here we have a moderate to severe high-frequency sloping loss.

This graph shows the hearing aid output and the match to target at a level of 65.

There is a good match to target up until around the 3 kHz point.

Next, we need to inspect the output of the hearing aid - this is the dark red line - regardless of how well it matches to target for the frequency regions that correspond to the ManU-IRU stimuli.

The low-frequency stimulus has a peak bandwidth of 240 to 611 Hz.

The mid-frequency covers 1,084 to 1,717 and the high frequency spans from 2,828 to 4,468 Hz.

We are interested in how much audibility the hearing aid output provides, so let's focus on that area.

For each frequency, we can see that the hearing aid output is definitely greater than 10 dB above the patient's threshold.

Therefore, we could sensibly test any of the 65 dB ManU-IRU speech tokens, expecting to see a cortical response.

If however, the hearing aid output is not greater than 10dB above the threshold, then we should look at the hearing aid output REM traces at 75 dB and see if they show greater than 10 dB above threshold.

If yes, then we can start testing at 75 dB speech reference level instead.