What happens to saccades if the medial and lateral recti muscles are suddenly
weakened, by surgery? What if the animal is kept in the dark during recover
from the surgery?
Optican & Robinson, (1980)
measure movements of both eyes
patch the weak eye, then patch the good eye
let the weak eye practice, with the good eye patched
try same paradigm without (some of) the cerebellum
unchangable increase in gain; affect of the pulse-step mismatch.
what happens when both eyes practice?
Reading
Lance Optican & David A. Robinson, "Cerebellar-dependent adaptive
control of primate saccadic system," J. Neurophysiol 44:
1058-1076 (1980) web
link
RHSC2 �13.4.1 & 13.5
Review of Optican & Robinson
Conclusions based on results of paper above seem weak: too many side effects.
from Adaptive Mechanisms in Gaze Control ? QP477.5 / A33
And see papers by Hare and Snow from Invest Ophthal.
Optican & Robinson, Methods. "The horizontal recti muscles of one eye were weakened by tenectomy... The tendon and about 5mm of muscle were cut from both the medial and lateral recti muscles of one eye. The rest of the muscle was allowed to retract into the rear of the orbit. The eye coil, previously implanted, was by this time attached to the globe by scar tissue and was unaffected by this procedure. Immediately following this operation, movement of the eye in the horizontal plane was severely restricted. Over a period of days the muscles reattached to the globe and became strong enough to move the eye approximately 10� away from the midline. A measure of the effective muscle strength was obtained by comparing the movements of the operated and un-operated eyes...[Fig. 1]..."
Figure 1, column A has the weakened eye patched! Note normal saccade from normal eye. column B has the normal eye patched and the weak eye viewing, for the first time. Notice staircase!
After sugery, but before experience: Pulse-step mismatch
overshoot on weak eye response: Caused by "scar tissue upsetting the ratio
of viscosity to elasticity in orbit" ...by? they guess: "increased
elastic stiffness relative to the viscosity..." What's the time constant?
r/k! To a step input discharge D the eye movement is
therefore if tau is decreased the response moves more rapidly to the asymptotic
value of A, and therefore overshoots, as noted by O&R. The final position
will be MN/k, a smaller value. The real reason for overshoot?
Given how the weakened muscle inserts further back, you'd think the elastic stiffness would decrease, but consider the slope of the muscle stress strain curve: At the top of the curve the slope is shallow.
Looks like the gain is about 1/3 after the tenectomy.
If the time constant's smaller, does that result in saccadic overshoot? YES! But should it also result in a saccade with increased speed? or are pulse muscles (FG) less affected than step muscles (SO)...? Or is it that the integrator has somehow decreased its gain?
The time constant of post-saccadic drift: 40 msec, versus the expected 150 - 200 msec. p. 1071. What a big difference! factor of 4-5 change in r/k?
Notice how RHSC defines the pulse gain, and the step gain.
Results
After experience of 5 days with normal eye covered, the P/S mismatch is corrected AND the saccade is of correct amplitude! See Fig 3. How long does it take the change to occur? Perhaps a day or less!
After the normal eye was patched the weak eye gain increased to 1, and the normal eye gain went above 2.0. Maximum gain faces a nonlinear saturation, it seems.
Total cerebellectomy
After total cerebellectomy, (what about DCN? see below for partial: fastigal nuc removed)
saccades can still be made permanent hypermetria...exactly compensates! see
Fig. 4, page 1067.
plus more drift, this time undershoot. recall the pulse being integrated to generate the
step. ...implies step gain > pulse gain ...2.9 vs 2.5... page 1066
saccadic oscillations for gains > 2... think about it, for gain of 1.5, we
end up 0.5 on the other side of the target, and the next saccade can get us
a little closer...
after cerebellectomy: neither size nor shape of saccade could be altered by experience.
RHSC p. 260: after cerebellectomy: spontaneous nystagmus: it greatly interferes with smooth pursuit.
Partial cerebellectomy
After partial cerebellectomy
(cerebellar vermis (medial zone), ...so deep that the fastigial nucleus was removed...)
(compare to cerebellar flocculus in the plastic VOR experiments)
wait less than a week, 5-6 days...
let weak, or normal eye have experience, not both...
Experience eliminates pulse-step mismatch, but
not hyper-size of saccade...Vermis lesion abolishes adaptive
control of pulse duration part of saccadic control.
FIGURE 6. weak evidence...In fact, where is the experiment of patching the weak eye and getting the normal eye to try to shrink its response back to normal?
O&R p. 1073: is the flocculus (=lateral region of cerebellum) concerned with step-gaze-holding ?? ... the flocculus reduces the brain-stem 15% over-size of step generation; with flocculus gone, we get too much step and "drift" back to target... and presumably the vermis is concerned with the pulse generation gain...how does this fit with the model from the saccade lecture?
Overall weak evidence that cerebellum controls plasticity directly, because the operated eye makes a "correct" saccade after cerebellum removed, and only pulse wrong after vermis...
Discussion
Separate control of pulse and step parts of the saccade. Why is the time constant of postsaccadic drift shorter (40 msec vs 200 msec for the eye in the orbit) ? mechanical change.
The equation D = k q + r qdot announces a relationship between
D and theta. D is not muscle force, but in normal operation is tightly linked
to force.
Now, once the muscles have been weakened and the relationship
between D and F changed, the equation becomes D* = k q + r qdot. If, in the
beginning, you just monitor the yoked control movement of the weakened patched
eye, you find a smaller waveform with overshoot. First consider the k term.
The weakened eye ends up at a smaller qa, (theta after). So one way to look
at it after:
D* = ka � qa + ra � dot-qa.
In fact, to begin with, D* = Db, D-before. So what is ka, or k after? Looks like D/qa = ka and ka > k. So far, this is just symbol manipulation for D*; in fact the actual ka may be less because the antagonist muscle is weakened too. At any rate, the weakened muscle, without visual experience, results in a new, smaller qa final steady state position.
What about r qdot the viscosity term? The weakened eye will go at a **slower speed** according to figure 1 data in the paper. Again this could be interpreted that ra is now larger than before, but in fact you'd think the surgery would have little effect on the actual r of the eyeball. What does not change for D* is the duration of the pulse of the command.
Back to the result of overshoot: What waveform do you expect if actual k is for some reason
lower and actual r remains the same? I think you'd expect that the lower force created by D*
in muscle will result in a ka qa that is greater proprtionally than the speed X time ramp the
pulse will get from a weakened muscle. Think about the situation. The result should be
undershoot, if that were true! How can we tell what's going on with k and r after? One way:
look at the time constant of the overshoot return. What is the time constant of the equation?
q (s) (k + sr) = D(s) so
and the time constant is r/k. If the
time constant is shorter after, that means that r is proportionally less than k after. Robinson
saw this and suggested that scar tissue increased elastic coefficient (could have reduced
drag of the eye). Who can argue with the facts?
Let's continue to look at the waveform, now after experience and recalibaration. No more overshoot. and the speed is normal! What does that mean? it means that the control of pulse and step are separate, that the compensation needed to overcome elastic and viscous forces are separate operations.
Look also at the good eye at the right of fig. 1: now it shows undershoot! does that make sense? for a singular control system of saccades? Now check experiment where vermis is removed. saccade has proper waveform but is remains too large. Doe the vermis of the cerebellum help determine the duration of the pulse? Does the flocculus control the step size? Another way to look at it, for O & R: the flocculus corrects post saccadic drift.
paper by Optican (1985)
"...intrinsic delays of saccadic and smooth pursuit systems
are 250 and 130 msec respectively..."
consider the viscous and spring components of the oculomotor load, in the equation
"The control of saccade amplitude seems to be dependent upon midline cerebellar
structures (vermis and fastigial nucleus) while the suppression of post-saccadic
drift seems to depend on the cerebellar flocculus and paraflocculus."
Optican, p. 71, 313? in Adaptive Processes... QP474/A33
implies the step integrator in cerebellum? what stops the pulse just in time?
what if the pulse is OK and the step requires "feedback"?
do large saccades fall intentionally short, in order for a brief comparison to be made?
pulse step again...can be fooled by slip conditioning
in same book, Goldberg, p329. Supp. Coll.-evoked saccades do not change size, even
after "slip conditioning" page332
Long term vision with a weakened eye
Snow, Hare & Villis, Invest. Ophthalmal & Vis Sci. 26: 924-931 (1985)
Methods: cut tendons to lateral and medial recti of one eye.
If both eyes are allowed vision, then eventually (>4 weeks) binocular compensation
occurs...After "normalization" the good eye's saccades had lower peak velocities!
means that cerebellum can overcome yoked control? or improve it...
[1989 thought: try removing flocculus alone...]
no removal of cerebellum here. they also tested VOR gain
Leigh & Zee on the cerebellum: 16 entries in their index.
pp 215 - 218
p. 77: total cerebellectomy virtually abolishes optokinetic nystagmus and smooth pursuit.
Discussion of oculomotor adjustments from RHSC2, chpt 13
RHSC2 �13.2.2 Nonlinearities p. 378-79.
"An old observation, confirmend in 1941 but apparently not reexamined since, is that
transposition of pairs of eye muscles is followed by a remarkably rapid process of recovery
of functional coordination" [Am. J. Ophthalmol. 24: 1115] "...simultaneous transposition of
medial and inferior recti, and lateral and superior, led to a return of proper cooperation after
only eight days, even when the animals had been maintained in the dark."
compare to Roger Sperry's experiments of eye rotation in frog.
RHSC2 �13.4.1 Saccades p. 388-390.
"The system appears to be incapable of making differential adjustments in cases where the
muscles of one eye are weaker than the other."
p. 397: "Total cerebellectomy in the primate abolishes the adjustments to unilateral muscles weakness when the good eye is covered, described earlier, and indeed creates post-saccadic drift; after lesion confined to the vermis paravermis, and fastigial nuclei, saccades are accurate in amplitude but still show drift (O&R 80). This suggests that the adjustments of the pulse component that determines saccadic amplitude are regulated essentially by the vermis and fastigial nucleus, whereas the step component that corrects post-saccadic drift is associated with some other cerebellar area, very likely the flocculus (Optican, 1985)."
cerebellum--- What does the flocculus do? VOR adjustment? Check Lisberger's theory. and see RHSC2, page 359; Weak discussion on how the results from partial cerebellectomy differ from the results for total cerebellectomy.
DEMO: Prisms and saccades
Summary
* Saccades after weakened eye muscle
Methods
size change & pulse-step mismatch
* Role of visual experience (one eye or two...)
Yoked eye movements: both eyes use the same control signals.
* Role of cerebellum vermis vs flocculus
* Total cerebellectomy: no smooth pursuit, no OKN!
* A role for negative feedback?
* Saccades & Ballistic arm movements
* Vergence eye movements
accommodation vs disparity
...plus [slower] saccades...
Random archive notes
mention visual motor reaching: ballistic
vs guided reaching
first, don't watch the hand, then...watch the hand = feedback
demo with prisms
reaching in the dark...C. Harris page 95-
p. 103 "As Figure 6 shows, this rapid adaptation has nothing to do with
vision and a lot to do with the adapted arm."
p. 104 "Thus, although the adaptation is specific to the adapted arm, it
is not merely a change in the arm's motor programs."
What else might change about the arm? Its position sense
Reaching to a fixation point doesn't
involve saccades,
Juggling while wearing prisms does involve saccades
[With UG Jude Cannon (1994).
Make ballistic eye movements to sudden targets.
See correction to errors over several trials.
Try pointing to sounds.
Try juggling with prisms
Is ballistic reaching a motor saccade?]
consider a weakened arm muscle moved in the dark. it will have no such problem as the weakened eye muscle. proprioception from the elbow will tell it where it's at...