Vestibular contributions to visual stability

1
Vestibular contributions to visual stability

• Ronald Kaptein Jan van Gisbergen

Colloquium MBFYS, 7 november 2005
2
Visual stability
Introduction
maintaining a roughly veridical percept of
allocentric visual orientations despite changes
in head orientation.
3
Visual stability
Introduction

• Different sources of information
• Visual
• Somatosensory
• Auditory
• Proprioceptive
• Vestibular

4
Visual stability
Introduction visual stability
5
Visual stability
Introduction visual stability
?
1
2
6
Subjective visual vertical
Introduction - SVV
Sudden transition at large tilt
7
Subjective visual vertical
Introduction - SVV
Errors in subjective visual vertical
Errors in subjective body tilt
-cw -ccw
8
Subjective visual vertical
Introduction - SVV
9
Subjective visual vertical
Introduction - SVV
Kaptein Van Gisbergen, J Neurophysiol, 2004
Kaptein Van Gisbergen, J Neurophysiol, 2005
10
Vestibular processing
Introduction
1
2
11
Vestibular system
Introduction
Canals Otoliths
12
Semicircular canals
Introduction
Limitations poor response to constant-velocity
and low-frequency rotations (i.e a high-pass
filter)
13
Otoliths
Introduction
Limitations cannot discriminate between tilt and
translation (ambiguity problem)
14
Otoliths
Introduction
Ambiguity problem

• Neural strategies for otolith disambiguation
• Frequency segregation model
• Canal-otolith interaction model

15
Frequency-segregation model
Introduction
Based on the constant nature of gravity and the
transient nature of acceleration
16
Canal-otolith interaction model
Introduction
Head tilt leads to a canal signal, head
acceleration does not
17
Questions
Introduction

• How good is visual stability during head
  rotations in the dark?
• What is the role of canal and otolith signals in
  this process?
• How can the processing of canal and otolith
  signals be modeled?

18
METHODS
Methods Task 1
19
Vestibular rotation
Methods
Supine canals only
Upright canalsotoliths
G
Sinusoidal rotation Amplitude 15 Frequencies
0.05, 0.1, 0.2 0.4 Hz
20
TASK 1
Results
21
Task 1
Methods
While rotating, subjects judged the peak-peak
sway of various luminous lines which counter
rotated relative to the head, at different
amplitudes.
22
Task 1
Methods
Not enough counter rotation
Too much counter rotation
23
Task 1
Methods

• Updating gain (G) the amount of counter rotation
  necessary for perceptual spatial stability,
  expressed as a fraction of head-rotation
  amplitude.
• G0 No updating (Head-fixed line is perceived
  as stable in space)
• G1 Perfect updating

24
RESULTS 1
Methods Task 1
25
Raw data task 1
Results Task 1
1 subject, upright
26
Updating gain
Results Task 1
perfect updating
no updating
27
DISCUSSION 1
Discussion Task 1
28
Interpretation task 1
Discussion Task 1
2
29
Interpretation task 1
Discussion Task 1
30
Interpretation task 1
Discussion Task 1
31
Otolith canal contributions
Discussion Task 1
updating gain
otolithscanals
canals
32
Otolith canal contributions
Discussion Task 1
improvement in upright, due to gravity, is
low-pass
canals
otoliths
33
Can current models explain our results?
Discussion Task 1
Not straightforward both models predict low-pass
characteristics in upright condition.
canal-otolith interaction
frequency segregation
34
Linear-summation model for rotational updating
Discussion Task 1
35
Linear-summation model
Discussion Task 1
Interaction model
Filter model
36
Fits of linear-summation model
Discussion Task 1
Interaction model
Filter model
upright
upright
supine
supine
R2adj0.72
R2adj0.82
37
TASK 2
Methods Task 2
38
Task 2
Methods Task 2
While rotating, subjects judged the side-to-side
displacement of various LEDs which were stable
relative to the head or stable in space.
39
Task 2
Methods Task 2
40
Task 2
Methods Task 2
Updating gain (G) the amount of counter rotation
necessary for perceptual spatial stability,
expressed as a fraction of head-rotation
amplitude. Perceived translation (T) the
perceived spatial displacement of an LED situated
on the rotation axis.
41
RESULTS 2
Results Task 2
42
Raw data task 2
Results Task 2
1 subject, upright
43
Updating gain
Results Task 2
perfect updating
no updating
44
Perceived translation
Results Task 2
45
DISCUSSION 2
Discussion Task 2
46
GIF Resolution
Discussion Task 2
47
Further processing necessary
Discussion Task 2
48
Translation predictionsusing perfect integration
Discussion Task 2
Frequency segregation
Canal-otolith interaction
49
Translation predictionsusing leaky integration
Discussion Task 2
50
CONCLUSIONS
Conclusion
51
Conclusions
Conclusion

• Q How good is visual stability during head
  rotations in the dark?
• A
• Compensation for rotation is only partial but
  better for higher frequencies.
• Small illusionary translation percepts in
  upright condition at highest frequencies.

52
Conclusions
Conclusion

• Q What is the role of canal and otolith signals
  in maintaining visual stability?
• A
• Both otoliths and canals contribute to
  rotational updating.
• Illusionary translation percept is otolith driven

53
Conclusions
Conclusion

• Q How can the processing of canal and otolith
  signals be modeled?
• A
• Rotation Linear summation of canal and otolith
  cues.
• Translation Double leaky integration of
  internal estimate of acceleration.
• We are not yet able to discriminate between the
  two disambiguation schemes

54
Questions?
Conclusion