4 ms·

Eyes see trouble coming before brain notices

9 points by soundsop 17y ago

rw 17y ago
Indicates that reverse-engineering human intelligence is harder than merely understanding the brain.
- jonsen 17y ago
  I believe it's common to regard the eyes as extensions of the brain.
mkelly 17y ago
I wonder these cells do when they detect an object approaching. Is this the reflex for closing the eyelids, or something else?
- a-priori 17y ago
  This article doesn't get into the behavioural response to this, because they're solely concerned with characterizing things at the cellular level. Personally, I doubt this would cause the eyelids to close (stimulation of the eyelashes already does this). It may activate the startle reflex, or perhaps initiate a saccade to focus the eyes on the stimuli. More research will surely follow to figure out exactly what behavioural role this information plays. It's already known, for example, that cells in V1 (primary visual cortex) respond to similar motion. Either the V1 cells receive input from these ganglion cells, or their function is somehow complementary to this (a different time scale, for example). Also interesting, this new finding directly contradicts existing research: "It is generally assumed that there is no sizable proportion of motion detectors in the primate retina." (Bach and Hoffman, 1999; http://dx.doi.org/10.1016/S0042-6989(00)00106-1 http://dx.doi.org/10.1016/S0042-6989(00)00106-1)
jacquesm 17y ago
Reflexes are also interesting, apparently the round-trip time to the brain for information is so long that in some situations something quicker is needed. So, you could probably conclude that some thresholding, computation and feedback capability is present in all neural tissue, including the spinal cord. Apparently an octopus has one 'main' brain and 8 'auxiliary' brains, one for each of its tentacles.
- a-priori 17y ago
  Well, first of all octopi are invertebrates, so their nervous systems are quite dramatically different from ours. If I remember right, an octopus' axons are unmyelinated (I don't know much about invertebrates; I know some invertebrates have myelination, but I don't think octopi do). This means that it's quite a bit slower to transmit signals around in their bodies, so in order to react quickly they need bundles of neurons near the sources of the signals. Perhaps as a throw-back to our invertebrate ancestors, we still have unmyelinated fibres that transmit some kinds of information, such as some forms of pain. That's why if you stub your toe, you feel two waves of pain: one sharp and quick, and the other much duller. The second one travels over unmyelinated fibres, and so takes much longer to arrive in your brain. If all you had were unmyelinated fibres, and if control were centralized in your brain, you'd have a round-trip time of a few seconds to your feet. How could you control your legs for walking like this? Obviously, we do have myelinated axons, so things are quicker, but it's still impractical to involve the brain in most reactions, because it's large and complex and takes a while to reach a steady-state after unexpected stimuli.
  - jacquesm 17y ago
    Thank you, that was most interesting. I knew about the myelination (sp?), but did not realize that that affected the speed of transmission so much. Very interesting about the secondary system still being active, maybe that is what allows us to measure the distance of a stimulus to the brain by using the difference between the arrival times of the pulses ? Are you working in this field ?
    - a-priori 17y ago

9 points by soundsop 17y ago

rw 17y ago

Indicates that reverse-engineering human intelligence is harder than merely understanding the brain.

jonsen 17y ago
I believe it's common to regard the eyes as extensions of the brain.

mkelly 17y ago

I wonder these cells do when they detect an object approaching. Is this the reflex for closing the eyelids, or something else?

a-priori 17y ago
This article doesn't get into the behavioural response to this, because they're solely concerned with characterizing things at the cellular level. Personally, I doubt this would cause the eyelids to close (stimulation of the eyelashes already does this). It may activate the startle reflex, or perhaps initiate a saccade to focus the eyes on the stimuli. More research will surely follow to figure out exactly what behavioural role this information plays. It's already known, for example, that cells in V1 (primary visual cortex) respond to similar motion. Either the V1 cells receive input from these ganglion cells, or their function is somehow complementary to this (a different time scale, for example). Also interesting, this new finding directly contradicts existing research: "It is generally assumed that there is no sizable proportion of motion detectors in the primate retina." (Bach and Hoffman, 1999; http://dx.doi.org/10.1016/S0042-6989(00)00106-1 http://dx.doi.org/10.1016/S0042-6989(00)00106-1)

jacquesm 17y ago

Reflexes are also interesting, apparently the round-trip time to the brain for information is so long that in some situations something quicker is needed. So, you could probably conclude that some thresholding, computation and feedback capability is present in all neural tissue, including the spinal cord. Apparently an octopus has one 'main' brain and 8 'auxiliary' brains, one for each of its tentacles.

a-priori 17y ago
Well, first of all octopi are invertebrates, so their nervous systems are quite dramatically different from ours. If I remember right, an octopus' axons are unmyelinated (I don't know much about invertebrates; I know some invertebrates have myelination, but I don't think octopi do). This means that it's quite a bit slower to transmit signals around in their bodies, so in order to react quickly they need bundles of neurons near the sources of the signals. Perhaps as a throw-back to our invertebrate ancestors, we still have unmyelinated fibres that transmit some kinds of information, such as some forms of pain. That's why if you stub your toe, you feel two waves of pain: one sharp and quick, and the other much duller. The second one travels over unmyelinated fibres, and so takes much longer to arrive in your brain. If all you had were unmyelinated fibres, and if control were centralized in your brain, you'd have a round-trip time of a few seconds to your feet. How could you control your legs for walking like this? Obviously, we do have myelinated axons, so things are quicker, but it's still impractical to involve the brain in most reactions, because it's large and complex and takes a while to reach a steady-state after unexpected stimuli.
- jacquesm 17y ago
  Thank you, that was most interesting. I knew about the myelination (sp?), but did not realize that that affected the speed of transmission so much. Very interesting about the secondary system still being active, maybe that is what allows us to measure the distance of a stimulus to the brain by using the difference between the arrival times of the pulses ? Are you working in this field ?
  - a-priori 17y ago