Digital Hearing Aids: Magic and Marketing
by Mark Ross, Ph.D.
Editor: I think most of you know of Mark Ross. He's a hearing aid
user who is also a Principal Investigator of the Rehabilitation
Engineering Research Center (RERC) on Hearing Enhancement. When he talks
about hearing loss, everyone should listen.
Here are his recent thoughts on digital hearing aids. This article is
reproduced with Dr. Ross' kind permission. For other articles by Dr.
Ross, please visit http://www.hearingresearch.org/ross.htm.
~~~~~~~~~~~~~~~~~~~
It seems to me, as I examine the marketing literature and
testimonials regarding digital hearing aids, that it's time for a bit of
corrective skepticism. I keep reading and hearing about how people are
just thrilled with their new "digital hearing aids", comments
by dispensers how they've had "better results with X digital
hearing aid" than any other since they've been dispensing, and
claims by industry that this new technology offers unprecedented
listening benefits. An increasing number of full-page ads in newspapers
and magazines, as well as commercials on the radio and the TV, are
currently extolling the superiority of digital hearing aids. People just
don't buy "hearing aids" any more; they purchase
"digital" hearing aids. To mix a few cliché's, the bandwagon
is rolling and nobody is commenting that the Emperor may not be very
well attired.
As applied to hearing aids, the term "digital" only refers
to a way of coding the acoustic input information, i.e. digital signal
processing (DSP). And, judging from an article that appeared in the
March 2000 issue of The Hearing Journal, this may be the only factor
that the many digital hearing aids have in common. In this article, Gus
Mueller interviewed representatives from seven leading hearing aid
companies, asking them about specific features that might be unique to
their product. The responses of the representatives - all very competent
and respected professionals - were very informative but also very
different.
While all the hearing aids they describe are digital, this begs the
question of where the presumed "digital" benefits lie. Are we,
in other words, to assume that there is some magic inhering in the fact
that all of these aids utilize DSP regardless of how they go about it?
Does digital processing per se produce superior speech perception
scores, independent of the diverse speech processing strategies and
architectures that different digital hearing aids employ? If so, this
would imply that all current digital hearing aids, regardless of
possible performance differences between them, would yield results
superior to that obtained by the best analog hearing aid. It also
assumes that the specific speech processing strategy used to modify
speech signals is basically irrelevant. Stated in this fashion, I do not
think that many of us would accept these assumptions.
Where "digital" hearing aids are clearly superior are in
the connotations of the term, in how beautifully it lends itself to
current marketing appeals. There is no doubt that the term has
resonance, associated as it is with modernity, the power of computers,
all the impressive technical developments of the 21st century. People
apply the term "digital" in a magical, almost reverential
manner, believing that they are describing the cutting edge of current
technology - and, of course, they are: digital hearing aids are truly
technological wonders. But the question still remains: Do all of these
hearing aids actually produce superior listening performance only by
virtue of the fact they employ DSP?
Presumably, it is in the creative ways that digital processing is
able to modify and deliver speech signals that produces listening
superiority. However, instead of clinical research evidence in support
this assumption, superiority is simply assumed on the basic of the
technical descriptions of the hearing aids. As one reads the
descriptions of their circuitry, it is difficult not to be impressed
with them, they all seem so logical and are described so appealingly.
Below, I've quoted or paraphrased just some of these descriptions and
terms applied to advertised digital hearing aids as found in just one
recent issue of a trade journal.
· Digital (or adaptive) feedback suppression
· Digital feedback management algorithm
· Noise reduction by differentiating between noise and speech
· Cochlear dynamics sound processing with 14 overlapping bands
· Digital cochlear dynamics
· Speech intensification system
· Speech enhancement algorithm
· Enhanced sound stabilizer
· Enhanced speech intelligibility through sensitive voice processing
· Loudness and dynamic range correction
· ConTrast, artificial intelligence system enhancing critical speech
sounds
· Elaborate (as quoted) input and output multi-channel compressions
system
· Adjustable kneepoints, ratios, and selectable time constants
· Six (or seven) band, three channel
· Three channel AGC-I with syllabic compression, with broadband AGC-O
· Digital processing calculates and processes a virtual sound match for
users
· Microphone noise reduction
· VoiceSync - a comprehensive system that manages speech
· Digital perception processing
· Fine scale noise canceler
· Adaptive digital audio zoom
· Auto select multi-memory system
· Lowest compression threshold
· Dynamic speech re-coding
In listing these terms, I do not mean to disparage or minimize the
very real technical accomplishments they represent. Clearly, some
first-rate minds and intense research efforts have been engaged in their
development. Some of the terms describe familiar concepts (e.g. bands,
channels, feedback suppression, compression parameters, etc.), but
others are less clear beyond the clear intention that the feature is
supposed to somehow improve speech intelligibility. Actually, not being
an electrical engineer, I don't expect to understand the specific design
features of advanced hearing aid circuitry. Nor am I overly disturbed if
I can't follow the detailed psychoacoustic rationale of how the system
is supposed to work. What I am able to understand, however, are clinical
research projects, studies that compare the actual listening performance
of hearing-impaired people with and without specific features. In other
words, clinical validation. This is what is sorely lacking.
Technical descriptions, no matter how impressive they sound (pun
intended) are not a substitute for clinical studies. As far as I can
see, there has been almost no research published in peer-reviewed
journals on most of the features unique to digital hearing aids.
Instead, what appears to be happening is that the description and
complexity of the specific electroacoustic features are themselves being
interpreted as evidence of their performance effectiveness. But
technical sophistication and creative designs do not, by themselves,
translate into improved performance. This has to be directly examined.
Indeed, there is the very real possibility that hearing aids are being
designed that exceed the capacity of an impaired auditory system to
benefit from them. That the many "bells and whistles" now
being incorporated in many digital hearing aids are simply adding to the
cost, but do not result in improved listening performance.
I am not suggesting that we ignore the detailed testimonials from
satisfied clients, or the accumulating clinical experiences of
dispensers. These are clearly relevant and very important
considerations. I am asserting, however, that these are not enough.
Surely, dispensers also require a body of objective research, studies
that can help them guide their specific choices regarding a speech
processing strategy. Depending upon the subjective impressions of
clients, while valuable and clearly necessary, is not in itself
sufficient. Dispensers can hardly lay claim to being a scientifically
based discipline if only subjective judgements were used in selecting a
particular hearing aid or speech processing strategy.
My memory in this field goes sufficiently far back for me to recall
the many testimonials made about earlier generations of hearing aids.
Every new development has had its champions and successes compared to
previous models or technology. How can one possibly compare
superlatives? How can one, for example, accept the glowing
recommendations of a particular product by one dispenser, while at the
same time receiving similar glowing descriptions of a different product
by another dispenser? Everybody seems to have their "favorite"
fitting, the one they've had their "best luck" with, the one
resulting in the "most satisfied" clients they've ever had.
Clearly, everybody can't be right. If they're all right, then no one is,
since this implies little or no difference between the results that can
be obtained with all of the various favorites.
But basically, what I object to is the almost magical manner in which
the term "digital" is invoked, as if this itself ensured
superiority. We know, from the very creative "hype" studies
conducted by Ruth Bentler and her colleagues at the University of Iowa,
just how powerful the "power" of suggestion is. When clients
were told that one aid was digital and other was analog, they almost
always preferred the digital aid - even though the exact same digital
hearing aid was involved in both trials. The subjects, and even in at
least one case report a spouse, reported that they heard better with the
"digital" hearing aid. And perhaps they did - for a while.
Modern medicine recognizes the powerful mind-body interactions that
can occur and may even depend upon it under some circumstances. None of
us - dispensers and consumers alike - are immune from the power of
suggestion. If people believe that they are being "treated"
with the most advanced "medicine" that modern technology has
to offer, and if dispensers reinforce this with similar, sincere beliefs
of their own, then positive judgements from consumers should be no
surprise. But how far can the power of suggestion take someone? Will the
reality of their objective hearing capabilities forever remain in the
glow of the term "digital?" Will questions regarding the
relative value of different DSP algorithms (and different digital
hearing aids) forever remain muted?
I do not question the fact that digital hearing aids are quality
hearing instruments. And I also do not doubt some types of amplification
strategies are only possible with DSP and may indeed be beneficial.
Where I have my personal questions and doubts is, first, the paucity of
objective evidence regarding specific speech processing strategies and,
second, how the term itself is invoked as a substitute for this
evidence. Hearing aid dispensers should be selling demonstrably better
hearing, and all this implies, and not just depend upon marketing
"hype" to sell their hearing aids for them.
~~~~~~~~~~~~~~~~
Editor: for some interesting thoughts on this
article, please see Rick Ledbetter's response.
