Choose a brand that is the best one for YOU.
What is a cochlear implant? Imagine you’ve bought yourself a new computer. The computer hardware is the implanted part of the cochlear implant (AB’s HiRes 90k, MedEl’s Pulsar & Sonata, Cochlear’s Freedom), the computer processor is the external speech processor (AB’s Harmony, MedEl’s Opus or Cochlear’s Nucleus 5), and the software is the programme which is adjusted through a series of mappings with the audiologist (AB’s HiRes Fidelity 120, MedEl’s FineHearing, Cochlear’s SmartSound). The cochlear implant companies bring out upgrades in the form of additional software to improve hearing functionality – such as AB’s ClearVoice which reduces background noise to provide greater speech clarity and ease of listening in noisy environments.
Which one is the best? My cochlear implant team told me they were all the same. But are they? Remember, they are audiologists and surgeons, not electrical engineers. Plus, they’re not deaf, so they don’t have a personal interest in ensuring you are fitted with the best cochlear implant for you. Each brand has positives and negatives, in the implant electronics (which is the ultimate limiting factor in the stim that can be applied), the electrode array itself, the processor hardware, and the DSP software. The electronics in the implanted part are what you own; and any ability for processor hardware and software improvements on the outside for a better stim hinge on what the internal circuits can accurately yet quickly decode. Basically, it boils down to this: the processor is something replaced/upgraded every 3-5 years, and you do indeed need to live with it until the next upgrade comes out. The implanted electronics is something you’ll be living with for the next 20 years (or longer). Whatever level of technology in the stim you get when it’s switched on is what you’ll have to live with.
By a good margin, AB and MedEl have the best implant electronics with separate current sources for each electrode button. AB’s can fire 90k pulses per second on its 16 buttons simultaneously with both + .AND. – charges; MedEl can fire about 56k pulses (but simultaneously either – .OR. +) per second on its 12 electrodes; while Cochlear’s can only fire 30k pulses per second, and only one pulse at a time. This means AB has 128 virtual electrodes, MedEl has somewhere around 90 (but they don’t claim a number due to their triangular waveform)… while Cochlear only has 22, with an unapproved (in US) capacity of 43 with current shorting.
The physical electrode array itself is where the action is, especially if residual hearing is desired to be preserved. Here, MedEl is on top, while Cochlear is also rather good. That being said:
Preservation of residual hearing also greatly depends on the skill of the surgeon, with an overall average of about 65-70%. The best cochlear implant surgeon in the world is Rick Case at Washington University in St Louis (WUSTL) who has a 98% success rate, and John Niparko at Johns Hopkins University in Baltimore (JHU) is right with him.
Preservation of residual hearing is also subject to a third variable, besides manufacturer and surgeon: whether the electrode is placed perimodionally or floating. Perimodional (wrapped in a tight spiral around the core) placement puts the buttons closer to the nerves for a better stim with lower power and tighter focus (especially if current steering is enabled), and it also enables a deeper placement for a given length (tighter radius); but it also entails a higher risk to to residual hearing.
The processor hardware is important, as it entails both the number-crunching capacity of the chips needed for speech processing and digital noise reduction, and also the physical container:
On the processor capacity itself, all three are adequate for the stimulations that the implant electronics can deliver at present. That being said, the implant electronics for Cochlear is already stretched to capacity; while Dan estimates MedEl’s still has 5-7 years to go and AB’s at least a decade.
The physical packaging for the new Freedom 5 is rather good, with waterproofing technology borrowed from RION’s HB-54 series hearing aids; and being shrunk almost as small as the MedEl Opus2. With 50% of cochlear implants going into children (at least in the USA), Advanced Bionics’ relatively clumsy processor has turned off many parents and cochlear implant centres for infants and small children.
The DSP (Digital Signal Processing) software running in the processor is also important, as there are two main parts to it: the manipulation of the incoming sound across the frequency spectrum for noise reduction - or lack thereof for music - and the conversion of the processed sound into one of the various stimulation strategies (CIS, n-of-m (ACE), Hi-Res Parallel, Hi-Res Sequential, etc)… which is ultimately limited by the amount of memory and processing power of the electronics in the implant itself.
Audiologist Jo-Ann has programmed Advanced Bionics and Cochlear patients. The biggest thing she wishes recipients to know is that everyone’s brain is different. You cannot predict how you will do, so having a choice of processing or coding strategies is key. Just like with hearing aids, some people prefer the sound of Phonak some Oticon, etc. With AB, they have MPS, CIS, HiRes-S, HiRes-P, HiRes-P with 120, and HiRes-S with 120.
Audiologist Joanna has mostly dealt with Medel, 30% with Cochlear and 1 with AB. She says the implant itself is only part of the solution. Most of your work starts when the processor goes on and most of it is positive mental attitude coupled with lots of listening practice. The more you practice the better the discimination becomes. For some, clarity happens straight away – for others it takes time, so there is no easy answer.
Some people’s brains prefer the sound of HiRes-S vs HiRes-P because they stimulate differently. With Cochlear you really only have one choice. Cochlear patients cannot try out other strategies because there is only one power source. This is also what gives the audiologist flexibility in making a good program for you. Even if Cochlear has new software coming out; with one power source you are limited. Jo-Ann compares it to building a house with 22 rooms. If you only have one light switch you can turn all the lights on or all of the lights off. You need multiple power sources to turn off/on on each “room” individually. This is also why Cochlear rates are slower and they can get away with disposable batteries.
It’s always a great idea to contact each company and speak with one of their audiologists-ask as many questions as possible especially about the research and about the internal capabilities. It’s also great to find a centre that has done equal amounts of each company so they have a less biased opinion. The company audiologists are the ones that will be supporting your programming audiologist and helping your centre after you are implanted. Surgeons really only see you once a year and don’t know about the programming. Also they won’t know about the external returns or breakdowns. Another test for companies is to call their customer service line and see how fast you get hooked up to a real person. (Courtesy Jo-Ann, audiologist)
Check out AB Harmony v Nuke 5
AB’s is by far the leader in the implant electronics, with full current steering for 120 virtual electrodes from their 16 electrode buttons and HiRes speed delivering up to 90 thousand updates per second for the best stims, especially for music & for speech intelligibility in noise.
The downside to AB is in about 5% of implant surgeries because they only make two electrode arrays — The “antenna” that goes into your cochlea. Both the Helix, for perimodiolar placement, and 1J semi-curved, for lateral placement, work well… But not for everyone.
One of AB’s implantees, Mike Marzalek, is writing stimulation software to achieve better music enjoyment with his cochlear implant. It only runs on his special hand built speech processor. If successful, this will open the door to more people getting cochlear implants as you won’t have to be quite as deaf to get real benefit, not to mention faster & better aural rehab, which is often overlooked. Dan says,
While you’re on Mike’s site, be sure to view his biography page. Read the 6th paragraph; and if it doesn’t sink in, rinse and repeat until it does.
There are two other “CI Insiders” who are also guiding me; but what Mike has up his sleeve writing his own stim and testing it on himself as he goes along is the reason why I’ve chosen an Advanced Bionics implant.
This is in addition to AB’s ClearVoice noise reduction technology, now released in the UK (March 2010) and Canada (February 2010); but still awaiting FDA approval for release in the United States. That, coupled with Phonak’s buying of AB in September 2009 means that their 50 years of “front end” audio processing will be migrating to the next generation of BTE processors to be married up with the “back end,” the actual encoding of the signal for electrical hearing.
AB’s offerings are quite good; but they lack in one critical area: The electrode array itself. When it comes to preserving residual hearing, their older 1J semi-curved electrode in a lateral (hugging the outside wall) placement isn’t too bad, even though it was designed to be used with the positioner (which stopped being included after they resumed shipping c.2002). On the other hand, AB’s Helix array, which is designed for a perimodional (hugging the inside perimeter, close to the spiral ganglion) placement, makes hash as it’s threaded in.
Worse for AB is that AB’s chief electrode engineer, Janusz Kusma, was fired by CEO Jeff Greiner in a pissing contest, putting them years behind MedEl and even Cochlear when it comes to low insertion trauma.
Still worse, AB has no other offerings, such as compressed electrode arrays for partially-formed or ossified cochleas: You either get a 1J or a Helix array… Period.
AB’s processor is the largest of the three — about the size of the MedEl Opus 1.
MedEl is about 7 years behind AB on the electrode itself, with about 55,000 updates/second; but they only have partial current steering with their FineHearing; and an estimated 90 virtual electrodes with their triangular waveform. With the MedEl Sonata implant you’ll have 5-10 years of new stims still coming. Looking at size, the MedEl Opus2 processor is much smaller than the Nuke5.
The upside to MedEl over AB is their selection of electrodes for almost every type of cochlea imaginable; and custom built electrodes that nobody else can do.
There is quite a bit of antipathy about MedEl. The issue is in the surgical placement of the electrode into the cochlea itself; and although there are similarities, there are also wide differences, especially when attempting a perimodional (vs lateral) electrode placement when threading it into place. What you want, especially for the tricky perimodional placement while trying to preserve residual hearing, is a surgeon with a lot of experience with that cochlear implant brand. MedEl’s processor and implant electronics are almost as good as AB’s; and their variety of electrode arrays is as good as, maybe even better than, Cochlear’s.
Bringing up the rear (from the implant electronics technology, working outwards) is Cochlear. They are hopelessly buried by competitor AB and MedEl patents. Because of this, even though they have a slick new Nucleus 5 BTE w/remote, they are still limited to the same 30,000 updates per second as they were over a decade ago. And, even though they have 22 electrode buttons, the best they can do is 43 virtual electrodes with current shorting — and they still haven’t released it yet!
The Nuke is about the same as MedEl’s CI from 2006. Cochlear has almost nothing new. They have a huge marketing budget though so beware, not all that glitters is gold….
There is one area where the Nuke 5 is worth a very long, careful look: children. Even though the implant electronics are obsolete, the implant itself is quite robust (or at least previous Nucleus implants are), and with infants and small kids they can’t tell you if it’s not working properly. The Freedom 3G processor has a fingernail-sized LCD status display on it; while the Nuke has a dual purpose (and bidirectional) remote control that also is a status indicator, duplicating the processor’s LED status lights. This is handy because a parent or teacher can simply walk up to the child with the remote, push a button, and get instant feedback from the processors and implants. This can be helpful when the parents themselves aren’t diligent in watching their children’s hardware.
Best hearing performance – Only recently have independent comparative studies been published, and AB came out on top in all of them.
Best reliability – All manufacturers claim +99% reliability for the implant, but only AB includes the external processor in the calculation. AB doesn’t make a big deal about water resistance, but their warranty does cover water damage. Cochlear makes claims about water resistance, but their warranty does not cover water damage.
Best upgradeability – AB’s HiRes 90k uses about 25% of its capabilities. Cochlear’s implant is pretty much tapped out, while Med-El’s is somewhere in the middle. The implant is the part you plan to keep for a very long time.
The T-mic – MedEl and Cochlear have their mikes on top of the ear, like BTE hearing aids. But AB has the mike at the tip of the ear hook, which is right at the entrance to the ear canal. This means your ear shapes the sound normally. Most users find this works much better than the BTE mike (which is also included, and is selectable in software). And you can use phones and headphones just like a hearing person. It’s patented, so the other manufacturers can’t put the mike there.
Miniaturization of processors – AB was purchased by Sonovus, the holding company of Phonak, in January 2010. This means that AB is the only company with access to the super miniaturization technology and years of experience of a hearing aid company. And Phonak is among the best. While AB’s processor isn’t currently the smallest, look for excellent features down the road.
Spahr et al EH 2007 compares performance of cochlear implants from Advanced Bionics, Med-El and Cochlear.
Audiology Online: “Preliminary Comparison of Performance between Patients Fit with the CII Bionic Ear and Patients Fit with the Nucleus 3G.” M. Dorman (Arizona State University), A. Spahr (Arizona State University) and K. Kirk (Indiana University School of Medicine), June 2003
In an independent study supported by NIDCD grant and contributions from all cochlear implant manufacturers, researchers compare performance between adults who use HiResolution sound and those who use the Nucleus 3 System. Results from the study’s first reported findings indicate that there “are differences in performance between patients using the two implant systems.” “Significant differences in performance (p <>
With thanks to (1) Surgeons Dr Lavy and Prof. Saeed, speech therapist Liz Stott (2) Dan Schwartz, a deaf electrical engineer formerly in the hearing healthcare business – he is about to get a CI. (2) Jo-Ann, audiologist (4) Joanna Nicholson, audiologist (4) Howard Samuels, BEA Mentor (5) Advanced Bionics, Med-El, Cochlear.