I came across this little slideshow on the interwebs (as ya do) and found it extremely interesting. It is from the Scientific American Frontiers show, and demonstrates in both audio and correlating images what a person with a cochlear implant will hear.
Amazing to think that a human ear has 30,000 neurons and a cochlear implant (which uses electrical impulses to stimulate the auditory nerve) currently has 24 electrodes! (The slideshow says 22 — it is clearly not the most recent data but gives a useful representation anyway.)
Obviously, because I don’t hear, the associated audio means nothing to me — I can’t tell what the quality is like or how easily distinguishable the words are. But the accompanying images, with their varying degrees of pixellation, gave me (as a future cochlear implant recipient) an interesting and valuable insight into the relative quality of what is heard with cochlear implants.
The audio goes from static white noise to garbled voices to seemingly clear electronic voices. Its like you are focusing in on the sound. Audio is equally as stunning as the visual.
30,000 neurons is like x number s transistors, the 24 electrodes is like the number of hair cilia in the ear. IIRC (it’s been a long time)
Thanks Gary! 24 cilia in the ear is not a lot, is it?! Amazing how well it seems to work.
Apparently the whole issue with people in my situation is that we have significantly less cilia than normal (or even none at all), thus the electrodes do the job of the cilia and stimulate the auditory nerve. (It seems strange that even though it’s known as “nerve deafness”, the nerve itself is actually ok.)
It’s quite incredible, the results that can be achieved with these electronic implants!
Hi Vicki,
I read your post a couple of days ago but have only just had the chance to listen to the slideshow.
The 22 channel slide was quite impressive.
I have been struggling with how to say the next bit but here goes:
(Assuming it was a fairly accurate representation)
It sounded like a natural recording that had been altered to sound more modulated and synthetic as opposed to a recording that had been built up from nothing — or a completely synthetic computer voice. I have had worse results during mobile-phone calls.
The words were distinct and understandable and the amazing thing was: I could look at the photo and say yes, that sounds like Alan Alda (caveat: I knew it was him so I am not sure how much of that was my brain filling in the missing details)
This goes way beyond just understanding words and is getting into the territory of picking out individual voices. Bloody fantastic technology!