Saturday 27 October 2012

Article on iPhone and hearing aid integration from 'the hearing aid blog'


Speculation started several weeks ago at Apple’s WWDC that the new API’s in iOS 6 would create some kind of linkage between the iPhone and hearing aids. This speculation was ramped up by this July 19th AppleInsider article revealing two key patent applications from January 2011. This article will examine these patent applications individually; and later we’ll provide our own SWAG (Silly Wild-Assed Guess) to add fuel to the fire.

First, a bit of history…

Users of hearing aids received their first big break in 1947 when the late Sam Lybarger, the Father of the Modern Hearing Aid, accidentally developed the telephone coil:¹ For the first time, deaf and hard-of-hearing people could hold the phone receiver over their (hot!) vacuum tube-powered hearing aid, switch on the T-coil, and through inductive magnetic coupling the baseband audio was picked up, amplified, and then transmitted to the button earphones… And it worked. However, two things conspired to break telephone compatibility: First, in the early 1960′s as the market shifted away from body aids and into transistorized behind-the-ear (BTE) instruments, the Electrical Engineers at the European hearing aid manufacturers seemed to have forgotten Maxwell’s Equations, namely that that the induced EMF in the telecoil magnetic pickup coil is proportional to (œ) the scalar dot product of the flux line vectors, which are oriented on the axis of the two coils, i.e. the EMF v, the vectors A & B, the absolute value of the vector magnitudes |A| & |B|, and the angle θ between A & B:
v  œ A • B = |A| * |B| * cosine (θ)
As a brief reminder from high school trig, cos (0°) = 1 & cos (90°) = 0
Cosine function as it applies to the scalar dot product of the angle between the axes of a hearing aid telecoil and the voice coil of a telephone receiver or room induction "hearing" loop
Cosine function as it applies to the scalar dot product of the angle between the axes of a hearing aid telecoil and the voice coil of a telephone receiver or room induction “hearing” loop
Click to enlarge
[Note: The actual value of electromotive force v  is a pretty complex calculation using Maxwell's Equations; but the important thing to note is the constituent relationship between the induced telecoil audio signal and the axes of the coil orientations.]
Although like a squirrel stumbling upon a nut, this vertical alignment of the hearing aid telecoil in fact works with horizontal room induction “hearing” loops on the floor or in the ceiling, in fact it broke telephone compatibility — Nice going, guys.
The second factor that conspired against hearing aid telecoil efficacy was that back in the late 1970′s, the Bell System and their Western Electric telephone manufacturing subsidiary came out with a new, lower cost receiver that was more efficient — In fact, it was a bit too efficient, as it did a better job of confining the magnetic flux to the voice coil — magnet gap area… And hence weakening by over 10dB the induced voltage in the hearing aid telecoil, which was already weakened to almost zero by the vertical alignment by the European BTE manufacturers. In fact, some American ITE manufacturers — notably Starkey, Telex, and Argosy — had Engineers who understood Maxwell’s Equations, and used either pancake coils glued to the faceplate, or long thin telecoils on a ferrite core stuffed down into the canal portion of the shell.
From A Look at the Telecoil — Its Development and Potential, by David A. Preves, the longtime Chief Engineer at Starkey Labs
After a huge outcry, a powerful consumer movement was started In Washington, D.C. by a lobbyist named David Saks and his organization — OUT  (Organization for Use of the Telephone) — to ensure that persons with hearing loss would be able to use their hearing aids with the telephone.
In 1982, the federal government passed the Telecommunications for the Disabled Act which required that telephones be labeled as to their hearing aid compatibility. The resulting legislation from the Federal Communications Commission (FCC) brought engineers from hearing aid companies and telephone companies together for the first time to work on the compatibility problem of telephones and hearing aids. The outcome of these meetings in the early 1980s was a new measurement standard for determining whether a particular telephone was compatible for coupling with hearing aids via induction pickup. A telephone that produced the proper amount of magnetic leakage In the proper direction, as specified in the standard, could be labelled and sold as “Hearing-Aid Compatible.” The law specified that coin-operated telephones in public places like airports were required to be hearing aid compatible. These hearing aid-compatible telephones were identified by a blue grommet at the junction of the cord and the telephone handset.
This consumer-driven movement on behalf of people with hearing loss went much further In 1989 when all telephones with cords sold in the United States were required to be hearing-aid-compatible, and in 1991, when all cordless telephones sold in the United States were required to be hearing aid compatible.
TIA-1083 LogoFast forward to 2007, and the Telecommunications Industry Association had introduced the TIA-1083 logo program for mass-produced phones: Just go to your local WalMart or Target and look for a box with this logo:


Moving on to mobile phone connectivity…

Getting a mobile phone to work directly with hearing aids has been a recurring nightmare, and more so on the AT&T, T-Mobile (Deutche Telekom), and European 2G & 3G GSM networks, and somewhat less so on the North American Sprint and Verizon CDMA networks. As it turns out, although the mobile phones work in the UHF portion of the radio spectrum, they “burst” their data differently, with GSM bursting at several hundred packets per second — Right smack in the audio range. This causes two distinct problems: First, with older hearing aids, one would hear a terrible screeching sound whether the microphone or telephone pickup coil was used. This was caused by inadequate shielding, and more importantly a lack of RF bypass around the semiconductor junctions: What would happen is that this strong RF signal would be rectified by the p-n semiconductor junctions, with the burst envelopes in the audio range being demodulated just like an AM signal would be… And then amplified to full power and faithfully reproduced loudly screeched by the output amplifier stage straight into the user’s ears.
About a decade ago, mobile phones started to incorporate IEEE 802.15.4 “Bluetooth” wireless Personal Area Network (PAN) connectivity both for synching to a user’s desktop PC, and for connecting to the ubiquitous headset (the ones that look like the wearer has a cockroach on their ear). You’ll notice that we spell out “IEEE 802.15.4″ instead of using the more generic “Bluetooth” for two distinct reasons:
• To emphasize that, in general, the IEEE 802.15 family is similar to TCP/IP in general, and moreso to IEEE 802.11 “WiFi” in that it is a two-way protocol, i.e. that the transmitting station sends a packet of data along with error correcting codes and a checksum, and then the receiving station decodes the packet, verifies and corrects what errors it can, and then transmit back an ACK(nowledgement) signal. If the sending station does not receive an ACK, then it will send the packet again. This presents issues with power consumption and up to 150 mSec latency, which will be discussed below;
• To separate out the commonly used 802.15.4 Personal Area Network (PAN) standard that we all know from the still-evolving 802.15.6 Body Area Network (BAN) standard that we believe Apple may be implementing in iOS 6.
Let’s look at how “Bluetooth” is currently implemented with hearing aids for connectivity, and the significant drawbacks.
First and foremost, we need to understand that any digital reception in a hearing aid is going to consume extra power — And lots of it, due to the decoding operation. Add to this the 802.15.4 overhead of transmitting ACK signalling, even occasionally in A2DP (Advanced Audio Distribution Protocol), and it makes for a real issue. Austin-based Audiotoniq has mostly sidestepped this with their hearing aids by using a Li-ION cell; but in fact their hearing aid wireless communications protocol is only really for phone use and not continuous streaming (though they have a clever workaround for it).
Most every other manufacturer uses a “Bluetooth streamer,” which acts as a “relay station” communicating via 802.15.4 to the phone or other Bluetooth -equipped device, and then using a second transmitter to broadcast a proprietary Hearing Instrument Body Area Network (HI-BAN) signal to the hearing aids. There are three basic ways this is accomplished, and it’s important to understand the distinction, as it is key in understanding what we speculate Apple will be doing:
• Widex and Phonak use a 28 meter (10.6 mHz) “near field” digital signal.² Phonak and Widex also use 10.6 mHz for ear-to-ear communication between the instruments for binaural coordination of directional microphone beam steering, compression to maintain binaural localization, and also program shift. Widex also uses it for binaural “Phone Plus” operation and Phonak for CROS and BiCROS communications; and both manufacturers also use it for wireless programming;
• Starkey uses a 33 cm (900 mHz) UHF digital signal for streaming and ear-to-ear communications; however they also have direct-to-instrument broadcasting through their SurfLink Media transmitter, i.e. unlike the Widex TV-Dex media transmitter, no additional relay is used. However, Starkey also just released their SurfLink Mobile device, which can be used as a Bluetooth relay, and also as a remote mic up to 20 feet away — But it’s on backorder until at least fall 2012 due to unanticipated demand;
• GN ReSound uses a variation of a 2.4 gHz 802.15.4 signal — An “unofficial” 802.15.6 HI-BAN — for direct, low (under 10 mSec) latency, direct-to-instrument broadcasting from various Unite accessories to their Alera series hearing aids, as well as for remote control and wireless programming (with inter-ear coordination available 4Q2012). It is this style of direct-to-hearing aid broadcasting that we believe Apple will be implementing in software in iOS 6, by essentially “hacking” the 802.15.4 Bluetooth stack and turning it into a de facto 802.15.6 HI-BAN stack for low latency broadcasting.³

 If this direct-to-instrument 802.15.6 2.4 gHz digital broadcasting standard is indeed brought out by Apple forcing the Big Six hearing aid manufacturers as well as chipmakers such as Intricon, ON Semi and others to agree on a single standard, we at The Hearing Blog cannot overstate the significance of this to those of us in the hearing impaired community, as well as to hearing care professionals and sound reinforcement engineers.

Here are just several reasons why this will vastly improve the life of us in the hearing impaired community — And not just those who use an iPhone with hearing aids:
• This will elegantly solve the issue of people needing to carry or even wear a Bluetooth streamer relay to use their mobile phones, wirelessly bringing the audio into both ears;
• This will allow for all hearing aid users to have a very effective and inexpensive alternative to FM assistive devices (ALD’s), as what ReSound is now doing with their very good Unite Mini Mic will be duplicated by other HA manufacturers. We cannot understate both the efficacy and cost aspects of this approach, especially with pediatric hearing aid (and eventually CI) users, as current 72 mHz, 168 mHz (H band) and 216 mHz (N band) analog narrowband FM (6F3 NBFM) systems are plagued with interference, as well as high current drain for headworn devices, messy and unreliable direct audio input (DAI) cables, and troublesome neckloops which are subject to head movement drop-outs and electromagnetic interference;
• This will open the door for hearing aid and sound reinforcement manufacturers to use inexpensive, off-the-shelf chips for direct-to-hearing aid room-sized broadcasting from entertainment devices such as TV’s, stereos, and game consoles;
• Most excitingly, at least for this writer, is this will enable inexpensive wide area direct-to-instrument broadcasting in large venues such as airports, lecture halls, theaters, arenas, and mass transit using an open source standard. What’s exciting about this is that there is already in place mass-produced high power data transmission chips that can easily be adapted via firmware to implement 802.15.6 broadcasting, and in fact there’s a good probability you’ve received this very page via IEEE 802.11 WiFi — That’s right: The technology to deliver the broadcast signal to a wide area is already being mass produced, and all it will take is a firmware update and an analog-to-digital converter (ADC) to turn a $39 WiFi-enabled Linksys router into a transmitter that covers up to several hundred yards.
• Once in place, gone will be the close call we are currently on the edge of experiencing with obsolete technology “lock-in” of baseband induction “hearing loops,” which are being forced upon those of us in the hearing impaired community by those with no technical background such HLAA’s  David Myers and Brenda Battat, and worse by Juliette Sterkens and Janice Schacter, who don’t even have to “eat their own dog food” because they don’t have hearing losses themselves. The problems with magnetic flux line alignment causing orientation problems with telephones (which is, umm, why it’s called a “telecoil”) has been documented above; while the very real problem of electromagnetic interference (EMI) that cannot be filtered out (because of the very nature of the baseband beast) is well documented, and is plainly evident to those who either are forced to use it or have an actual knowledge of electromagnetic and communication engineering.
We already know that Apple is in close consultation with the Big Six hearing aid manufacturers: Given their current state of haphazard connecting Apple (and hence other) mobile phones to their hearing aids, this is the most likely initial part of the roadmap ahead. This brings us to…

But just what about those two patents breathlessly cited in the AppleInsider article?

Glad you asked! Let’s look at them individually at first and then together, in the context of the S.W.A.G. we just laid out.
• Remotely updating a hearing aid profile, United States Patent Application 20120183165 (PDF here)
• Social network for sharing a hearing aid setting, United States Patent Application 20120183164 (PDF here)
Although at first blush these patents look sexy, let’s look at them individually:
Remotely updating a hearing aid program (or for that matter, cochlear implant MAP) is something that has already been done by America Hears (and their partner Australia Hears, now Blamey & Saunders) for over a decade, and what Audiotoniq is using through mobile handsets. In fact, there is a possibility that Apple’s patent application for this function is invalid, as it represents prior art and is henceforth not patentable.
Sharing hearing aid settings through GPS-based “Foursquare” social networking is indeed a possibility; however there are significant HIPAA (privacy) issues in play. However, this would still involve communications between the iPhone and hearing aids… Via 802.15.6, as described above.

UPDATE: Near-Field Communications will .NOT. be supported on the iPhone 5

We originally penned this article eight weeks ago; but held off because of the near-field communications (NFC) wild card Apple could have played. Unlike the software changes outlined above in iOS 6, NFC requires an additional hardware chip. What we didn’t know until the iPhone 5 release two days ago, the “Made for Apple” hearing aids will also work on the iPhone 4s — Which would have made it obvious that it did not involve the NFC protocol.
For more on what NFC is and why Apple did not include it, please see iPhone 5 NFC snub explained by Apple in c|net UK.

References & Footnotes:

2) This use of 10.6 mHz presents a problem for Phonak’s Advanced Bionics division’s CI’s, as their Clarion II and HiRes 90k implants also use a very weak 10.6 mHz signal for the reverse updates to monitor implant integrity and telemetry, the interference of it causing a loss-of-lock and instantaneous shutdown of the implant circuit. This means that “bimodal” (CI + HA) users cannot avail themselves of the iCom or other wireless 10.6mHz technology
3) There is also the possibility that Apple has plans for using the 802.15.6 standard for a more general body-area network for connecting other medical devices such as pulse monitors, blood glucose monitors, and other things; but the FDA Device Branch will have the final say~