Interview with David Speidel, Director of Audiology Services, Interacoustics
Share:
Topic: How Interacoustics Has Improved ASSR Testing
Paul Dybala: Hello everybody. This is Dr. Paul Dybala with Audiology Online, and today I'm talking with David Speidel from Interacoustics (interacoustics-us.com). He is the Director of Audiology Services there, and he's here today to talk to us about their new ASSR testing system. David, thanks for taking time out to visit with me.
David Speidel: You're welcome.
Dybala: Give us a little bit of background just on yourself and Interacoustics.
Speidel: Sure. I started this newly developed position back in November of 2006. At that time we realized as a company, that as we were getting into higher-end instrumentation and driven by PCs that there was a growing need for additional customer support.
"Support" can be looked at in a variety of ways. The obvious is phone support; but another is in a forum like this, publishing our articles, doing courses for Audiology Online and things of this nature.
We are ramping up our support services here in the company so we can better support the people who purchase our equipment. It also gives us an ongoing opportunity to educate our distributors, who are ultimately the people who are in contact with the end users. So, we look at it as a growing area into which we want to put some effort and concentration.
Interacoustics has been around since the early 1970's. Traditionally we're known for making audiometric hardware, such as audiometers, impedance units and hearing instrument test equipment. We've been lesser known in the ABR area, although we have had some ABR instrumentation on the market for seven or eight years.
As we have grown and looked to the future, we have developed quite a nice research and development department that encompasses approximately 45-50 people just working on new hardware and new software.
Dybala: That's a good sized group.
Speidel: Yes, software is very time intensive. You don't realize how much effort it takes until you get into this business.
It takes a coordinated effort in the R&D department to look at how we can incorporate hardware and make it multifaceted, so that we can also use it for other applications. Today we are talking about ASSR today, but it really is an add-on to a product that we've already developed.
For example, the recently released Eclipse platform looks like a small medical device, which is connected to a computer. From there, we've added our ABR system, which is capable of doing electrocochleography, typical ABRs, middle latency, late latency, P300, and MMN (mismatch negativity) tests. So, it's really a comprehensive ABR system.
Now we've built on to that by adding otoacoustic emissions, both distortion product and transient evoked. So, now we've got ABR, OAE, electrocochleography, and now the addition of ASSR.
We look at this as a building process. The advantage, from an end-user standpoint is that the system is upgradeable. As each software interface is being developed, we integrate it so each application looks similar to the next. For customers who are buying our product now, it's just so much easier for them to learn the product because it looks so similar to something they've already had.
Dybala: Sure. So, if you've been performing ABRs using Interacoustics software and now want to do ASSRs, you will already be in that same software, with that same look and feel. Correct?
Speidel: That's correct.
Dybala: Just to get into the ASSR, it is a "relatively" new test in the field. Could you just give us an overview of actually what ASSR is?
Speidel: Well, ASSR stands for Auditory Steady-State Response and it is an offshoot of ABR testing. I think there's some confusion among audiologists of what ASSR actually is and where the responses are derived. It still remains a response that is generated by providing some type of a stimulus to the auditory system, therefore it's electrophysiologic in nature.
What we're trying to do is modify the stimulus in order to pick up on a response. It isn't like an ABR response where you see a wave I, III, and V, and you're trying to mark each wave. With ASSR, we're just looking for a statistical probability of the presence or absence of a response, whereas with the typical ABR, you're depending on a subjective analysis of the information by a trained eye to mark a wave.
Example Test Results from Interacoustics' ASSR Software
With the ASSR, an algorithm detects the probability of an electrical response, so you can say affirmatively that there was a response at a given decibel level. What makes ASSR different from the earlier systems is the delivery of the stimulus.
With ABR, we have always used a click or toneburst stimulus. With ASSR, you're basically modulating a steady tone, either by amplitude or frequency modulation, or a combination of the two. Based on the modulation, there is some type of electrical activity that occurs in the auditory system, and that is what we're trying to detect.
Dybala: What level of the auditory system is the ASSR actually looking at?
Speidel: That's a really good question, because it depends. For example, if you were to use a high modulation rate, somewhere in the neighborhood of 90 to 100 Hz, that actually would be an early brain stem response. This is typically the rate most people use. But there are systems out there now, including ours that are also capable of a 40 Hz modulation rate. This rate moves the locus up towards the cortex, so now you're looking at a response that's primarily being generated from a middle-latency region.
It's quite interesting. There are different reasons for using either one of those, but again we're not looking so much at a waveform as the probability of a response.
Dybala: And if I understand it correctly, ASSR is currently being used mostly for threshold estimation at multiple frequencies, correct?
Speidel: That is correct. What's interesting about the whole topic of ASSR or the technology is that you can actually subject the auditory system to more than one stimulus at a time by varying the modulation rate. As long as you separate the modulation rate of those frequencies by a little bit, you can generate a response at each of those frequencies. Thereby, you could test four frequencies in one ear at a time. With our system we allow you to do this binaurally as well.
Dybala: Tell me more about Interacoustics' new generation of ASSR testing and what makes your system different from other systems out there.
Speidel: That's a good question because the systems out there right now, as I mentioned, modulate either amplitude or a frequency. The system is typically looking at the modulation rate of a fundamental frequency.
So let's say, for example, if at 1,000 Hz, you're modulating at 90 Hz. The system's algorithms are trying to detect that 90 Hz response.
Researchers recently discovered that there are harmonics above the primary modulation rate. So, you could go from 90 Hz to 180 to 270 and so on, and look at the additional harmonics. Now, as we all know, if you can add more data points to your algorithm in your analysis, statistically what will happen?
Dybala: You could detect a response faster, I would assume.
Speidel: Yes, exactly. And so, we actually take into consideration and additional six or seven harmonics. The researchers who have studied this did a specific study where they looked at each individual component of the harmonics to find out the statistical contribution that they could bring to the overall result. They found that the all harmonics add a significant amount of information which would speed up the test time. Their analysis concluded that, by using this type of an algorithm, you could reduce your test time by as much as 50%.
Dybala: So, you're saying that your ASSR testing system is essentially 50% faster than other conventional systems, correct?
Speidel: That is correct.
Dybala: Well, that's a significant improvement, I would say. What else makes it different?
Speidel: Well, the other factor I mentioned is the stimulus. Scientists have creatively manipulated the stimulus to account for cochlear travel times. In general, as you know, with standard ABR it is difficult to obtain really nice data in the low frequencies at a rapid rate. That's due to the fact of cochlear travel time. If you stimulate with a 500 Hz stimulus, it takes much longer for it to travel down to the basilar membrane. The difference is fairly significant between a 4,000 Hz test carrier signal and a 500 Hz toneburst. The researchers were able to generate a stimulus that actually introduces the lower-frequency spectrum first.
Based on a cochlear travel time model, they designed the envelope so that the 500 Hz signals arrive at the same time on the basilar membrane as the 4,000 Hz signals. The idea here is that we're going to create a more robust signal and more synchronous neural firing, which is the key to any neural physiologic measurement. The outcome of this was that the response amplitude was much, much larger than what was currently being used.
Dybala: Interesting. It sounds like they're just very clever at understanding the properties of the signal and of the auditory system to take advantage of how those things interact.
Speidel: Exactly. I'm just amazed at what they've done here. But the other thing is that, in current ASSR systems, the stimulus itself has a certain envelope to it. You can imagine that it has a certain region where it gets to be very frequency-specific, which, as you decrease in intensity, what happens? It stimulates fewer and fewer hair cells along the basilar membrane. As the amplitude decreases, fewer hair cells are stimulated.
From here, researchers decided to broaden the stimulus a little bit. For example, rather than physically modulating the stimulus, they separated the cosines of the stimulus by 90 Hz. By including approximately 7 cosines spaced by 90Hz they created more of a narrowband type of stimulus. Each cosine being separated by 90 hertz. Well, what happens with this?
Dybala: Beating.
Speidel: Correct, so this beating or modulation is generated because of the way the stimulus is designed. In addition to that, because they widened the stimulus envelope of the cosines, they also increased the amplitude of each cosine so it would stimulate a larger section of the basilar membrane. This enables us to get a faster detection in low frequencies.
Dybala: Well, it sounds like researchers have done a lot of work in thinking this through.
Speidel: Yeah. Over the years I have looked at ASSR casually, of course, and have tried to understand it, even though we didn't have all of this in our bag of tricks.
It is so cool what they've done with this stimulus, not only introducing the lower frequencies first before the highs, but the way they designed it to create the modulation rate. Overall, of course, this has contributed to reduced test time. This is not only due to the algorithm, but also due to the type of stimulus that's being used, because now we can create a more robust signal and electrical response, and the detection is much easier to get.
Dybala: So it's a faster test; it's more robust. It sounds like it can be quite frequency-specific. Just one other question I wanted to ask before we sign off, David. Is there anything else new coming out from Interacoustics in the testing and evaluation areas?
Speidel: Yes. We've expanded our VNG system to incorporate what's called FireWire technology. This is related more to camera technology. But of course, when we're doing videonystagmography, we're using cameras, so it just refers to the transmission rate and our ability to have a higher resolution with tracking eye movements. That's going to be released the early part of 2008.
Dybala: We'll have to schedule another interview talk about that!
Speidel: That'd be great.
Dybala: I did want to mention that if anybody is interested in additional information on this new ASSR technology from Interacoustics, visit interacoustics-us.com and you can also head over to their Web Channel from our home page at Audiology Online, or by clicking to audiologyonline.com/channels/interacoustics.asp. They have live and recorded seminars on this and other topics. We typically offer that for CEUs, so not only can you learn more information about this great technology, you can earn some CEUs as well.
With that, David, I just want to say thanks again for your time.
Speidel: You're welcome, and thank you! If any of your readers have any questions at all, they can contact me via the Interacoustics website. I would be happy to share additional information.
David Speidel: You're welcome.
Dybala: Give us a little bit of background just on yourself and Interacoustics.
Speidel: Sure. I started this newly developed position back in November of 2006. At that time we realized as a company, that as we were getting into higher-end instrumentation and driven by PCs that there was a growing need for additional customer support.
"Support" can be looked at in a variety of ways. The obvious is phone support; but another is in a forum like this, publishing our articles, doing courses for Audiology Online and things of this nature.
We are ramping up our support services here in the company so we can better support the people who purchase our equipment. It also gives us an ongoing opportunity to educate our distributors, who are ultimately the people who are in contact with the end users. So, we look at it as a growing area into which we want to put some effort and concentration.
Interacoustics has been around since the early 1970's. Traditionally we're known for making audiometric hardware, such as audiometers, impedance units and hearing instrument test equipment. We've been lesser known in the ABR area, although we have had some ABR instrumentation on the market for seven or eight years.
As we have grown and looked to the future, we have developed quite a nice research and development department that encompasses approximately 45-50 people just working on new hardware and new software.
Dybala: That's a good sized group.
Speidel: Yes, software is very time intensive. You don't realize how much effort it takes until you get into this business.
It takes a coordinated effort in the R&D department to look at how we can incorporate hardware and make it multifaceted, so that we can also use it for other applications. Today we are talking about ASSR today, but it really is an add-on to a product that we've already developed.
For example, the recently released Eclipse platform looks like a small medical device, which is connected to a computer. From there, we've added our ABR system, which is capable of doing electrocochleography, typical ABRs, middle latency, late latency, P300, and MMN (mismatch negativity) tests. So, it's really a comprehensive ABR system.
Now we've built on to that by adding otoacoustic emissions, both distortion product and transient evoked. So, now we've got ABR, OAE, electrocochleography, and now the addition of ASSR.
We look at this as a building process. The advantage, from an end-user standpoint is that the system is upgradeable. As each software interface is being developed, we integrate it so each application looks similar to the next. For customers who are buying our product now, it's just so much easier for them to learn the product because it looks so similar to something they've already had.
Dybala: Sure. So, if you've been performing ABRs using Interacoustics software and now want to do ASSRs, you will already be in that same software, with that same look and feel. Correct?
Speidel: That's correct.
Dybala: Just to get into the ASSR, it is a "relatively" new test in the field. Could you just give us an overview of actually what ASSR is?
Speidel: Well, ASSR stands for Auditory Steady-State Response and it is an offshoot of ABR testing. I think there's some confusion among audiologists of what ASSR actually is and where the responses are derived. It still remains a response that is generated by providing some type of a stimulus to the auditory system, therefore it's electrophysiologic in nature.
What we're trying to do is modify the stimulus in order to pick up on a response. It isn't like an ABR response where you see a wave I, III, and V, and you're trying to mark each wave. With ASSR, we're just looking for a statistical probability of the presence or absence of a response, whereas with the typical ABR, you're depending on a subjective analysis of the information by a trained eye to mark a wave.
Example Test Results from Interacoustics' ASSR Software
With the ASSR, an algorithm detects the probability of an electrical response, so you can say affirmatively that there was a response at a given decibel level. What makes ASSR different from the earlier systems is the delivery of the stimulus.
With ABR, we have always used a click or toneburst stimulus. With ASSR, you're basically modulating a steady tone, either by amplitude or frequency modulation, or a combination of the two. Based on the modulation, there is some type of electrical activity that occurs in the auditory system, and that is what we're trying to detect.
Dybala: What level of the auditory system is the ASSR actually looking at?
Speidel: That's a really good question, because it depends. For example, if you were to use a high modulation rate, somewhere in the neighborhood of 90 to 100 Hz, that actually would be an early brain stem response. This is typically the rate most people use. But there are systems out there now, including ours that are also capable of a 40 Hz modulation rate. This rate moves the locus up towards the cortex, so now you're looking at a response that's primarily being generated from a middle-latency region.
It's quite interesting. There are different reasons for using either one of those, but again we're not looking so much at a waveform as the probability of a response.
Dybala: And if I understand it correctly, ASSR is currently being used mostly for threshold estimation at multiple frequencies, correct?
Speidel: That is correct. What's interesting about the whole topic of ASSR or the technology is that you can actually subject the auditory system to more than one stimulus at a time by varying the modulation rate. As long as you separate the modulation rate of those frequencies by a little bit, you can generate a response at each of those frequencies. Thereby, you could test four frequencies in one ear at a time. With our system we allow you to do this binaurally as well.
Dybala: Tell me more about Interacoustics' new generation of ASSR testing and what makes your system different from other systems out there.
Speidel: That's a good question because the systems out there right now, as I mentioned, modulate either amplitude or a frequency. The system is typically looking at the modulation rate of a fundamental frequency.
So let's say, for example, if at 1,000 Hz, you're modulating at 90 Hz. The system's algorithms are trying to detect that 90 Hz response.
Researchers recently discovered that there are harmonics above the primary modulation rate. So, you could go from 90 Hz to 180 to 270 and so on, and look at the additional harmonics. Now, as we all know, if you can add more data points to your algorithm in your analysis, statistically what will happen?
Dybala: You could detect a response faster, I would assume.
Speidel: Yes, exactly. And so, we actually take into consideration and additional six or seven harmonics. The researchers who have studied this did a specific study where they looked at each individual component of the harmonics to find out the statistical contribution that they could bring to the overall result. They found that the all harmonics add a significant amount of information which would speed up the test time. Their analysis concluded that, by using this type of an algorithm, you could reduce your test time by as much as 50%.
Dybala: So, you're saying that your ASSR testing system is essentially 50% faster than other conventional systems, correct?
Speidel: That is correct.
Dybala: Well, that's a significant improvement, I would say. What else makes it different?
Speidel: Well, the other factor I mentioned is the stimulus. Scientists have creatively manipulated the stimulus to account for cochlear travel times. In general, as you know, with standard ABR it is difficult to obtain really nice data in the low frequencies at a rapid rate. That's due to the fact of cochlear travel time. If you stimulate with a 500 Hz stimulus, it takes much longer for it to travel down to the basilar membrane. The difference is fairly significant between a 4,000 Hz test carrier signal and a 500 Hz toneburst. The researchers were able to generate a stimulus that actually introduces the lower-frequency spectrum first.
Based on a cochlear travel time model, they designed the envelope so that the 500 Hz signals arrive at the same time on the basilar membrane as the 4,000 Hz signals. The idea here is that we're going to create a more robust signal and more synchronous neural firing, which is the key to any neural physiologic measurement. The outcome of this was that the response amplitude was much, much larger than what was currently being used.
Dybala: Interesting. It sounds like they're just very clever at understanding the properties of the signal and of the auditory system to take advantage of how those things interact.
Speidel: Exactly. I'm just amazed at what they've done here. But the other thing is that, in current ASSR systems, the stimulus itself has a certain envelope to it. You can imagine that it has a certain region where it gets to be very frequency-specific, which, as you decrease in intensity, what happens? It stimulates fewer and fewer hair cells along the basilar membrane. As the amplitude decreases, fewer hair cells are stimulated.
From here, researchers decided to broaden the stimulus a little bit. For example, rather than physically modulating the stimulus, they separated the cosines of the stimulus by 90 Hz. By including approximately 7 cosines spaced by 90Hz they created more of a narrowband type of stimulus. Each cosine being separated by 90 hertz. Well, what happens with this?
Dybala: Beating.
Speidel: Correct, so this beating or modulation is generated because of the way the stimulus is designed. In addition to that, because they widened the stimulus envelope of the cosines, they also increased the amplitude of each cosine so it would stimulate a larger section of the basilar membrane. This enables us to get a faster detection in low frequencies.
Dybala: Well, it sounds like researchers have done a lot of work in thinking this through.
Speidel: Yeah. Over the years I have looked at ASSR casually, of course, and have tried to understand it, even though we didn't have all of this in our bag of tricks.
It is so cool what they've done with this stimulus, not only introducing the lower frequencies first before the highs, but the way they designed it to create the modulation rate. Overall, of course, this has contributed to reduced test time. This is not only due to the algorithm, but also due to the type of stimulus that's being used, because now we can create a more robust signal and electrical response, and the detection is much easier to get.
Dybala: So it's a faster test; it's more robust. It sounds like it can be quite frequency-specific. Just one other question I wanted to ask before we sign off, David. Is there anything else new coming out from Interacoustics in the testing and evaluation areas?
Speidel: Yes. We've expanded our VNG system to incorporate what's called FireWire technology. This is related more to camera technology. But of course, when we're doing videonystagmography, we're using cameras, so it just refers to the transmission rate and our ability to have a higher resolution with tracking eye movements. That's going to be released the early part of 2008.
Dybala: We'll have to schedule another interview talk about that!
Speidel: That'd be great.
Dybala: I did want to mention that if anybody is interested in additional information on this new ASSR technology from Interacoustics, visit interacoustics-us.com and you can also head over to their Web Channel from our home page at Audiology Online, or by clicking to audiologyonline.com/channels/interacoustics.asp. They have live and recorded seminars on this and other topics. We typically offer that for CEUs, so not only can you learn more information about this great technology, you can earn some CEUs as well.
With that, David, I just want to say thanks again for your time.
Speidel: You're welcome, and thank you! If any of your readers have any questions at all, they can contact me via the Interacoustics website. I would be happy to share additional information.
