The eardrums move when the eyes move up, down, and sideways: our latest on eye movement-related eardrum oscillations – what information they carry, and what they might indicate about how vision and hearing operate as partners. A tweeprint! 1/N biorxiv.org/content/10.110…
Proud of the team: David L. K. Murphy is lead author, with @cindyking40, Stephanie N. Schlebusch, Christopher A. Shera, and me. This line of research was originally initiated by former PhD student, now Capt. Dr. Kurtis Gruters. 2/
Background: your brain is the boss of your senses. It controls what the eyes see via blinks, focus, and eye movements. Hearing is a little different - we humans can’t close or move our ears - but our brains do have some tricks up their sleeves 3/
The interior parts of the ear contain moving parts you can think of as muscles – the middle ear muscles and specialized cells in the cochlea called outer hair cells 4/
These motor elements are controlled by the brain and linked to one another via the moveable parts they are attached to – the bones of the middle ear, the basilar membrane of the cochlea - via the mechanical coupling between these structures. 5/
Because of this coupling, when any of these parts move, they cause the eardrum to move 6/
When the eardrum moves, it makes a sound. Yup. We usually think of the eardrum as responding to sound, but it can also make a sound when these motor elements cause it to move 7/
These sounds are usually too quiet to hear, but you can measure them by placing a microphone in the ear canal 8/
Some of the sounds generated this way are called otoacoustic emissions (OAEs). They were discovered in 1978 by Kemp and are now routinely examined as part of clinical assessments of hearing 9/
Mostly, we’ve thought of this system as being useful for tamping down the ear’s responsiveness in loud settings, e.g. shutting down the sound of our own voices, as well as amplifying really quiet sounds. 10/
Two years ago, we discovered something surprising: the eardrum moves when the *eyes* move as well! 11/ pnas.org/content/115/6/…
We presume the brain is using these in-ear-motor mechanisms in an eye-movement-related way. But why would it do *that*? We suspected the answer has to do with calculations the brain has to perform to be able to link what we see and what we hear. 12/
The brain detects the locations of visual stimuli by noticing where on the retina the image falls. However, remember that the brain can move the eyes? And every time it does so, the image of a particular stimulus shifts on the retina 13/
Now suppose you’re out birdwatching. You want to match up the bird you see with the bird you hear. Your brain knows the locus of the bird’s image on the retina. How does it tell where the song is coming from? 14/
To localize a sound, your brain has to compare sound arrival time and loudness across the two ears. It also pays attention to the frequency content - because the folds of your ears filter frequency in a direction-dependent fashion. 15/
By themselves, these cues are insufficient to link with the visual system: they are using different frames of reference. Visual input is anchored to the moveable retina, whereas the auditory cues depend on the sound’s location with respect to the head and ears. 16/
When we saw that the eardrums were moving in connection with eye movements, we immediately wondered if this might mean there is an underlying mechanism that provides information about eye movements to the auditory periphery 17/
In our new study, we measured these signals in human subjects using a bunch of different tasks. Our basic idea is that if this mechanism is to be useful for linking vision and hearing…18/
… it should carry accurate information about horizontal and vertical eye movements, and information about both the change in eye position and the absolute eye position the eye movement is beginning or ending at. 19/
TL; DR – it does! And furthermore, these effects are reproducible enough that you might be able to eye track from microphone recordings within the ear canal. 20/
Caveat – not arguing this is a desirable way to eye track 🤷21/
The point is proof of principle about what information is present in this signal and what it means for the underlying, as-yet-unknown neural mechanisms 22/
Some additional points of note. 23/
We get asked frequently whether this could just be the sound of the eyes moving in the orbits. The fact that the eardrum signal slightly precedes the onset of the eye movement, and extends after it for tens of milliseconds, would seem to rule this out. 24/
We think this is likely a normal element of hearing and should not cause you any problems with localizing sounds should they occur during an eye movement. @LeanneBoucher Howard Hughes, and I looked at this previously 25/ sciencedirect.com/science/articl…
Finally we hope more people will join us in investigating this phenomenon! There is lots to do. We hope to publish a methods paper soon that will help audio-folks with eye tracking and oculomotor-folks with audio techniques. Comments, questions welcome! 26/26
Did I break the thread? Continues hère:
