Exploring the Audible Contrast Threshold (ACT) Test: From Science to Clinical Practice in Hearing Aid Fittings

Thank you for joining for

this presentation.

The ACT test unveiled a Game

changer for precise hearing

aid amplification in

noisy environments.

My name is Ashley Hughes and I’m

an audiologist at Interacoustics

US.

I began working at

Interacoustics just over

four years ago,

but have been practicing in a

variety of settings for

over ten years.

Prior to joining

Interacoustics,

I first worked clinically,

then I worked as a research

audiologist for a hearing aid

manufacturer and then I joined

Interacoustics in 2020.

In my current role at

Interacoustics, I provide support,

education,

and training on all things

audiometry, impedance OAEs,

hearing aid fitting,

verification, and telehealth.

And of course,

with that now comes support and

education on ACT more

to come on that.

As mentioned,

today’s focus will be

on the new ACT test,

which stands for Audible

Contrast Threshold developed

by Interacoustics.

During today’s session,

you’ll learn the requirements

and benefits to both the patient

and clinician of incorporating

and using the ACT test,

how to prepare for and complete

the ACT test, again,

both from the patient

perspective and the provider

perspective,

and finally,

how to interpret and utilize the

ACT value to improve patient

outcomes and provide value to

both the patient and

the provider.

To learn this,

we’ll also spend a fair amount

of time today reviewing the

research that led to the

development of actual

on this slide you can see a

brief overview of our

agenda for today.

Well start with discussing

why ACT was developed.

Audiometry is an important

clinical tool,

but there are aspects of hearing

that the pure tone audiogram

doesn’t cover.

Think about what information we

don’t get when performing

pure tone audiometry.

We get some information from

pure tone. Audiometry,

but not the whole picture.

The audiogram doesn’t tell us how

the patient will perform

in the real world.

The ability to hear pure tones

in quiet does not reflect the

typical demands that everyday

conversational situations

impose.

In using only.

Pure tone audiometry to fit

hearing aids is limited in the

guidance we then have for

advanced hearing aid

feature settings.

Finally,

as many studies have shown,

we know already that pure tone

audiometry is a poor predictor

of intelligibility.

Speech audiometry is something

that clinicians perform at times

to supplement audiometry.

Performing these tests provides

the clinician with

more information,

and we can see and understand

some of the reasons why speech

audiometry is beneficial.

It certainly has its place,

and there are lots of studies to

suggest speech in noise tests

specifically can help supplement

the patient rehabilitation

process.

Additionally,

speech audiometry provides

relatable material

for the patient.

It can help guide hearing

aid management,

it can be used during the

counseling process and more.

Now we can all agree why speech

testing can be very beneficial,

and there are plenty of articles

suggesting why it

should be done,

but what exactly prevents some

clinicians from doing it?

We certainly know that in real

life situations it can be

difficult to include speech

audiometry within our

appointments,

despite the advantages

of including it.

Some of these obstacles

are time,

access to testing materials,

objective and clear starting

points for clear audibility for

the patient room setup,

and more.

So,

we know that the audiogram

is vital,

it’s what’s mandatory for

hearing aid prescription

targets.

But we get no information about

speech and noise abilities from

the pure tone audiogram.

Speech audiometry is a great

additional to any consultation,

but does have its limitations

in some cases,

and not everyone is doing

it in any case.

So what marries the two?

What can give us accurate

diagnostic information that

takes us beyond the audiogram

that’s unique to each patient

and can address the number one

concern reported by patients

with hearing loss?

Hearing and noise.

And this is where act

comes into play.

On the next few slides,

we’re going to walk through the

typical patient workflow.

Based on today’s standards,

we’re typically starting

with an audiogram

then moving on to advanced

diagnostic tests when needed,

such as the TEN test.

Speech in quiet testing is

often performed SRT, Word rec,

MCl, UCL, etcetera.

And at times we’re moving on to

speech in noise tests like the

quicksand, ANL things like that.

These are great indicators for

hearing aid acceptance,

but they’re not completely

objective tests and can’t be

used automatically to adjust

advanced features and

hearing aids.

And after that point, we’re.

Course,

moving on to the hearing aid

fitting when appropriate.

So with the patient in mind,

a gold standard consultation

could look like incomes act,

the accurate predictor of aided

speech and noise performance,

independent of language and

easy to administer.

The purpose of this workflow is

to show what could be possible

you would still perform

audiometry.

It’s a baseline requirement

for ACT.

ACT isn’t replacing any part of

a clinician’s existing workflow

if they don’t want it to,

but it can be used as a

replacement for other speech

and noise tests.

It’s a fast test designed to be

used alongside your audiogram in

order to make a comprehensive

and customized fitting

for your patient.

Lets take a look back at the

research that has led to the

creation of ACT and the pathway

from research to diagnostic

implementation.

To best understand ACT,

were going to take a journey

from where it started

to where we are now.

Well go through everything

from the history of act,

who the brains are behind

the invention,

what is the background evidence

behind act that makes it

accurate and reliable?

And how did we create a user

friendly diagnostic

implementation so it can be

added to your existing workflow?

So why ACT and why now?

Because the only diagnostic

measurement we currently use to

program hearing aids

is the audiogram,

which is still extremely

valuable.

But as we’ve already discussed,

hearing loss is so much more

than the audiogram.

There has been an ongoing search

for diagnostic tests that take

us to the next level beyond

the audiogram,

but no one was able to

confidently and successfully

create this. Yet.

An act is about to change this.

If we think of the audiogram as

measuring the quantity

of the hearing,

act measures the quality

of the hearing.

We still have our audiogram

and real ear measurements,

of course,

that ensure the hearing aids are

programmed appropriately

for the hearing loss,

acoustic parameters of the

patient’s ears, and more.

But now we also have ACT,

another tool in our toolkit that

helps guide the programming of

advanced help and noise features

in hearing aids.

These two combined provide you

with the tools you need to get

a complete diagnosis of your

patient’s hearing and a more

objective and customized

fitting.

This has been a giant

undertaking and because of this,

I want to make sure to give

credit where it’s due.

To the collaborative team that

was involved in the

creation of ACT,

we have had interacoustics

research unit heading up seven

years of research.

The journey also involved inner

acoustics for diagnostic

implementation.

Eriklsholm research Center.

And CARR,

the Center for Applied

Audiology Research.

Together,

we developed ACT and transformed

it into a unique automatic

prescription.

More to come on this,

but for now,

we’ll move on to the research.

So where did it all begin?

It began in the 2010s,

when research saw a strong

correlation between a test of

something similar to ACT and

speech in noise measures.

This early version of ACT,

as we now call it,

is known as spectrotemporal

modulation, or STM,

and it’s a psychoacoustic

measure.

STM is essentially an

unmodulated sound with a

modulated sound imposed

on top of it.

As an example of what STM is,

we can see on the left a

spectrogram of a single

sentence.

As you know,

this shows characteristic

patterns across frequency

and time.

These patterns are the same as

in an early ACT or STM stimulus

shown on the right.

Modulations are imposed

onto a noise signal,

so you get a similar pattern as

the speech spectrogram

on the left,

but in a more controlled

and structured fashion.

Let’s hear some sound

examples of this.

First,

you’re going to hear

the ACT stimulus.

Next you’ll hear the reference

noise or the steady noise.

And finally,

you’ll hear the ACT stimulus

imposed on top of the

steady noise.

If you’re good at act,

you can detect tiny modulations,

which in essence means you’re

good at detecting the important

speech cues when embedded

in background noise.

By manipulating the depth of the

modulations and getting patients

to detect these modulations,

we can obtain a threshold or an

act value from which we can

infer how good someone will be

at picking out the speech

cues and noise.

However,

there were problems with this

early version of ACT.

A body of literature on STM was

studied and there was interest

gained in this in 2013,

where they were showing

correlations between early

versions of ACT against other

speech in noise measures.

The results from this

more controlled,

carefully recruited group

of test subjects showed

correlations were very high with

beyond the audiogram measures.

In 2016, a subsequent study,

our early act test,

was taken through a large

study in Sweden,

but one third of the population

couldn’t do the test.

They couldn’t hear

the modulations,

which was very frustrating

for that one third that failed.

They had to modify the test

paradigm so that they could

get a measurement,

which is where the data points

come from on the slide

at this point,

IRU really wanted to unleash the

potential of this test and make

it clinically viable as act

as we know it today.

So what modifications did IRU

have to make to this

early version of.

STM to make it clinically

viable.

As we’ve said,

earlier versions of

spectrotemporal modulation were

not easy tests to administer.

So let’s take a closer look of

what we had to do to unleash

its full potential.

The first change was to ensure

that the stimulus was

clearly audible,

to ensure that patients could

actually do the test.

In previous tests,

we talked about how there was a

ceiling effect where essentially

one third of the participants

couldn’t complete the test.

The first modification was to

add ear and frequency specific

compensation for hearing loss.

In previous STM tests,

in order to make the

stimuli audible,

they cranked the signal up,

but IRU took a different

approach.

On the slide you can see a graph

showing gain compensation.

Here’s a closer up image of that

graph as we go through it.

In the green line you can see

hearing thresholds of normal

hearing people.

In the black line you can see

the ACT stimulus played to

someone as if they have

normal hearing.

You can see that across

the frequency range,

it’s aligned with the natural

speech levels and everything

is super threshold.

Now,

if we take a look

at the red line,

we have an audiogram of an

individual with hearing loss.

If we were to present the black

stimulus to people with an

audiogram that looks

like a red line,

they wouldn’t hear the

entire stimulus.

The entire stimulus is

not super threshold.

And now you can see what Iru did

differently in the pink bars,

increasing the necessary

frequencies by 15 decibels

sensation level

to make them clearly audible and

therefore following the shape

of the audiogram,

IRU ensured that the stimulus

was clearly audible to fight

the sailing effect,

particularly in older

individuals due to cognition

etcetera.

They also changed the test

paradigm to a three alternative

forest choice paradigm in order

to make the differentiation

between stimuli easier.

In this test paradigm,

one sound of the three presented

was different and the test

subject was required to pick the

odd one out rather than

in previous studies,

they had to explain the

difference between two

presented stimuli.

During this iteration,

the modulation times were also

increased from 0.5 seconds

to 1 second.

Finally,

during these evaluations,

IRU decided to go with a

binaural stimulation instead

of a monorail stimulation,

increasing the ease and accuracy

of the test as well as improving

the correspondence with real

world speech and noise

listening situations.

Next,

they had to adjust the degree of

spectral modulation and the

bandwidth of the stimulus.

To do this,

the researchers needed to look

at other speech and

noise measures.

In collaboration with Eriksholm,

IRU set up a test environment

that imitated a real world

listening environment,

but using loudspeakers this

study looked at actual.

In relation to speech and noise

measures, in this case,

the hint.

The hint was used since it can

be set up in an ecologically

valid fashion against simulating

real world listening situations.

They found a high correlation

between act values

and hint scores,

and all study participants were

able to produce a threshold,

resolving the ceiling effect and

reinforcing the relationship

between act and speech and

noise performance.

To summarize,

what IRU completed

until this point,

they compensated for audibility,

ensuring super threshold

presentation.

They changed the test paradigm

to a three alternative

fours choice.

They extended the length

of the modulations,

they went from monaural to

a binaural presentation.

They ensured that the task and

stimuli parameters provide

reliable results across

the board,

and they ensured that the test

has a high correlation to speech

and noise testing performed in

an ecologically valid fashion.

At this point,

we knew that ACT works,

and we wanted to create a

clinically viable tool.

In order to do this,

IRU conducted further research

to ensure the test

time was quick.

There was no need for

extra hardware.

It used a familiar procedure

that audiologists are used to,

and the value it produces takes

into account the best available

data points for a more precise

threshold determination.

So, after all this research,

what makes ACT so special?

We have a foundation for

a better first fit.

The act value has a strong

correlation to predicted aided

performance in noisy

environments,

the most common concern reported

by patients with hearing loss.

This value can help guide you on

how to adjust advanced features

on the hearing aids.

At the first fitting for

select hearing aids,

the ACT value can be input into

the hearing aid manufacturer

software,

and the adaptive features of

that hearing aid can be

automatically adjusted.

It is the first ever language

independent diagnostic test of

speech in noise ability.

It’s conducted in a very similar

way to pure tone audiometry

using the modified Houston

Westlake bracketing approach,

but with two in four decibels

step sizes instead

of five and ten,

making it an easy test to

perform for both the patient

and clinician.

Other speech in noise tests can

take significantly more time

than ACT, which takes,

on average, two minutes.

And we have a novel way of

denoting the act value,

the nCL scale,

which was created in order for

us to understand how act values

compare to normal hearing

listeners.

Now let’s take a look at the act

software and how we perform it.

I’ll show you what it looks like

today using the interacoustics

Affinity compact.

This is a close up of the ACT

software screen. You can see.

The presentation numbers on the

x axis and the ACT value

in dBnCl on the y axis.

There’s a line at 25

presentations because this is

where you begin to risk

patient fatigue.

In order to avoid hitting or

going over 25 presentations,

it’s very important not to time

lock the patient during testing.

Channel one and channel two

presentation levels will be

predetermined based on the

audiogram and one channel will

be routed to each ear.

The test will always start

at 16 dB nCL,

which you can see in the top

center of the screen,

and as the nCL value decreases,

the modulation is more

difficult to detect.

On this next screen,

you can see a completed

ACT test.

When the clinician introduces

the modulation,

the software will automatically

determine if the patient

responded or not based on the

timing of the presentation

and the patient response.

A correct response will show as

a black filled in circle and

an outline circle is a miss

in the top right hand corner.

You’ll be able to see

false alarms.

Every time the patient

correctly responds,

the clinician will make it four

decibels more difficult

or two steps,

and when the patient misses it,

the clinician will make it two

decibels easier or one step

again following the modified

Houston Westlake bracketing

approach,

but using two and four decibels,

step sizes instead

of five and ten.

Note that the clinician is not

adjusting the intensity of

channels one or two,

but rather the difference

between the modulated signal

and the base noise.

When the patient responds

correctly at the same intensity

three of five times on

an ascending trial,

the software will automatically

stop you and provide you

with the ACT value.

Once you have the ACT value,

you can click on guidance to see

the ACT value ranges

and fitting advice.

We’ll take a look at those

in just a moment.

With all this information

on ACT,

we’ll spend a bit of time

talking about how you can go

about using the results.

Generally speaking,

the greater the ACT value,

the more help one will need from

the hearing aids in noise.

You can use the ACT value

in one of two ways.

If the hearing aid manufacturer

supports it,

you can use an automatic

function and the value of the

ACT will be transferred into the

hearing aid manufacturer

software to determine the level

that the adaptive feature

should be set at.

These changes can be made again

automatically through the

hearing aid manufacturer

software.

However, in some cases,

if that isn’t an option,

you can use the guidance

that we provide.

You can see from the scale that

the act value will fall into one

of four categories ranging

from normal to severe.

The lower the number,

the smaller of a modulation.

The patient was able to detect

again if somebody.

If someone has a lower

ACT value,

that means they’re able to

detect even smaller modulations.

So the recommendation for these

patients is to pull back on

those advanced help and

noise features.

Those patients are doing just

fine at detecting speech in

noise without extra help.

On the other hand,

with a higher ACT value,

you’ll want to do the opposite

and increase those advanced

features,

perhaps recommend some

additional accessories for that

patient from the very beginning.

If we think about fittings

as they exist today,

two patients with the same

audiogram, same age, etcetera,

would likely leave an office

with the same help in

noise settings.

With the introduction of ACT,

we can now be confident on when

we want to pull down on some of

those advanced adaptive features

and perhaps crank these features

up for some patients,

meaning two people of the same

age with the same audiogram will

leave the clinic with two

different and customized fitting

setups from the very beginning,

which in turn will improve

the first fitting.

In summary, we know that ACT,

or the audible contrast

threshold test,

is a clinically viable test of

spectrotemporal modulation

detection.

It uses the audiogram to

individually shape and customize

the stimulus to ensure full

audibility across the

frequency range.

ACT as a measure of an

individual’s aided predicted

speech in noise performance

using hearing aids in realistic

conditions.

It is a fast and simple test.

It takes, on average,

two minutes.

It can be done immediately

after the audiogram,

and it is a language

independent test,

meaning it can be done on every

patient irrespective of their

language background.

If a patient is able to follow

pure tone audiometry

instructions,

they’ll be able to follow

ACT instructions.

Expected outcomes when using

act are to provide you with

confidence when you’re

prescribing or adjusting those

advanced help and noise features

on hearing aids.

Improved patient outcomes

in noisy situations,

reduced follow up visits,

reduced in the drawer

hearing aids.

And we’re very excited about the

more research that we’ll be able

to share as it becomes available

on the ACT test.

Thank you for taking the

time to join today.

If you’re interested in learning

more about act and you’re

in the United states,

please contact your local

e3 Diagnostics sales

representative.

For those outside the US,

please reach out to

Interacoustic directly at

[email protected]

as shown on the slide.

Thank you.