An Agenda for Sensorimotor Research in Sub-Orbital Flight · Experienced Day 1 Day 2 Day 3 0 0.5 1...

An Agenda for Sensorimotor Research in Sub-Orbital Flight

1

Faisal Karmali, Harvard/MEEIMark Shelhamer, Johns Hopkins

Supported by: NSBRI, NIH, NSERC (Canada)

Special thanks: Ondrej Juhasz, Michelle Zwernemann, John Yaniec, Noel Skinner and our research subjects

2

How can we: Make sub-orbital flight more safe? Make sub-orbital flight more enjoyable? Benefit humans on Earth through

sub-orbital flight experiments?

Sensorimotor disruptions

3

AutonomicMotion Sickness

Orthostatic intolerance

Eye movements

Vestibulo-ocular reflexPitch

Roll

Eye alignmentTorsional

VerticalNystagmus

Saccade accuracy

Postural stabilityGait

Head-body-eye coordination

Motor coordination Manual joystick control

Orientation illusions Inversion illusion

Goals

4

Sub-orbital passengers

Sub-orbital pilots

• Maximize enjoyment (maximize corporate revenue via customer referrals)

• Accomplish research tasks

• Safely pilot the aircraft during both nominal and emergency situationsWithout interference from sensorimotor disruption

Overcoming sensorimotor disruptions

5

Strategies

Adaptation

Re-adaptation

Pre-adaptation

Pharmaceutical

Cognitive training

Orbital flight

Sub-orbital passengers

Sub-orbital pilots

***** * *

** * *****

*** ***** **

*** *** **

***** ***** ***

How quickly do we adapt?

6Experienced Pilots Day 1 Early Day 1 Late Day 3

-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

Naive

Diff

eren

ce b

etw

een

0 g

and

1.8

g in

pitc

h ve

stib

uloo

cula

r ref

lex

gain

(deg

/deg

)

Recommendations

7

Operators ResearchersUse pre-adaptation in parabolic flight for sub-orbital passengers

Emphasize recency of experience for sub-orbital pilots

Develop and conduct a neurological examination for sub-orbital pilots

Consider screening of passengers for latent and undiagnosed neurovestibular problems before sub-orbital flight

Study the effectiveness of parabolic flight as a tool to pre-adapt sub-orbital passengers

Find sensorimotor symptoms caused by the unique sub-orbital flight trajectory, and ways to mitigate them

Study the correlation between sensorimotor disruption and pilot performance and appropriate countermeasures.

Adaptation

Adaptation Seconds - Days

9

Research Goals

10

Sub-orbital passengers Sub-orbital pilots

• Allow passengers to fully focus on the flight experience without distraction of sensorimotor disruption

• Allow passengers to complete a set of personal tasks within a short period of time, such as movement in 0 g, flips, looking out the window, and interacting with other passengers

• Allow scientist-passengers to complete scientific tasks quickly and accurately

• Consider any interactions between flight phases specific to sub-orbital flight

• Ensure that sensorimotor disruption does not interfere with the ability to pilot the aircraft during both nominal and emergency situations

• Quantify re-adaptation capability by study the effect of gaps in sub-orbital exposure on functional neurological tests and actual flight performance metrics

11

Adaptation of otolith-ocular responses to parabolic flight

Presented by:Faisal KarmaliResearch FellowJVPL / MEEI / HMS

October 14, 2008

Supported by: NSBRI, NIH, NSERC (Canada)

Faisal Karmali, Ondrej Juhasz, Michelle Zwernemann, Mark Shelhamer

12

Take-home points In parabolic flight, certain otolith-dependent ocular

responses improve over the course of 3 days of flying

Pilots (non-parabolic) had similar responses to experienced parabolic fliers, suggesting that certain experiences can prepare one for parabolic flight

Adaptation of otolith-dependent pitch responses transfers to otolith-dependent translation responses

13Experienced Day 1 Day 2 Day 3

0

0.5

1

1.5

2

2.5In

stab

ility

of to

rsio

nal a

lignm

ent (

deg)

Naïve Subjects(n=5)

Pilot Subjects(n=3)

Experiencedparabolicflyers(n=4)

Naïve subjects had significantly differentresults compared to both pilots and experienced subjects on day 1 (p<0.04),but not subsequent days.


Pilot Subjects(n=3)




Pilot Subjects(n=3)



14None Single episode vomiting Severe vomiting

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Motion sickness score

Inst

abilit

y of

tors

iona

l alig

nmen

t (de

g)

Naive subjects (n=6)Pilots subjects (n=3)Experienced subjects (n=4)

15

Day 1

0 g 1 g 1.8 g0.5

0.6

0.7

0.8

0.9

1

1.1

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)

Naive subjects (n=3)Pilot subjects (n=2)

0 g 1 g 1.8 g0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)

Naive subjects (n=3)Pilot subjects (n=2)Experienced parabolic flyers (n=4)

Day 3

16

Take-home points In parabolic flight, certain otolith-dependent

neurological responses improve over the course of 3 days of flying

Pilots (non-parabolic) had similar responses to experienced parabolic fliers, suggesting that certain experiences can prepare one for parabolic flight

Adaptation of otolith-dependent pitch responses transfers to otolith-dependent translation responses

17

Measuring static eye position• OCR• Torsional alignment• Vertical alignment

• Nikon D70 digital camera

• Subjects’ head upright/ tilted

18

Measuring pitch VOR gain Active sine-like head

movements • 0.3-1.6 Hz. Results

based on 0.6-1.3 Hz• 20-30º• 40-90º/sec

Fixating stationary point in the light.

Eye movements were recorded using a head-mounted video system (Chronos).

Gain computed by least-squares fitting of eye velocity to head velocity, for each cycle of motion.

20

Adventures in weightlessness:

Adaptation of otolith-dependent ocular responses to parabolic flight

Presented by:Faisal KarmaliResearch FellowJVPL / MEEI / HMS

MEEI Vestibular SeminarMarch 24, 2008

Faisal Karmali, Ondrej Juhasz, Michelle Zwernemann, Mark Shelhamer

24

The bottom line…1. Parabolic flight consists of specific flight trajectories

that provide periods of 0 g and 1.8 g.

2. Subjects experience sensorimotor disruption when initially exposed to parabolic flight, specifically in the pitch vestibuloocular reflex and torsional alignment.

3. Over the course of three days in parabolic flight, responses became appropriate in 0 g, 1 g and 1.8 g.

4. Responses in 1 g were not affected by adaptation in 0 g and 1.8 g, suggesting context-specific adaptation (rather than generalized sensory rearrangement)

25

Outline Background: Parabolic flight Background: Science Methods: Recording eye position Results

Conclusions

26

BackgroundParabolic Flight

27Mercury astronauts training aboard a C-131B (1959)

28

Play movie: C:\Faisal\kc-135\Skew\KC135_March2002_Day3_MarkFaisal.avi

Play movie:C:\Faisal\kc-135\Pictures\Aug2006\walkingupsidedown.mov

29

Start of 0 gPitch 45º nose-up

24000 feet(7300 m)

34000 feet (10000 m)

0 g

1.8 g 1.8 g

End of 0 gPitch 45º nose-down

g = gia = gravitoinertial acceleration ⇒ the occupants’perceptions

of gravity

30

Play movie:c:\Faisal\kc-135\Pictures\Apr2006\HPIM2161.MPG

31

It is not zero gravity! In parabolic flight, the plane accelerates downwards to

match gravity, so the net force is zero. Even in orbit, there is still gravity (~9.37 m/s2 at 300 km)

freefall weightlessness microgravity zero gravityzero ggravitoinertial acceleration (gia)

Preferred terms Less correct

32

KC135 Photo

33

Why doesn’t everybody fall to the front of the plane?

Wz

WWx

Nz

Nx

Vertical (z)axis

Longitudinal (x) axis

θ

degreesKarmali F, Shelhamer M. The dynamics of parabolic flight: flight characteristics and passenger percepts. Acta Astronautica (In press)

34

0

0.5

1

1.5

2

Ver

tical

g le

vel (

g)

Aircraft g level (gravito-inertial acceleration) during two parabolas (aircraft coordinates)

0

0.5

1

Long

itudi

nal

g le

vel (

g)

0

0.5

1

Late

ral

g le

vel (

g)

0 20 40 60 80 100 120 140 160 180

Alti

tude

Time (seconds)

350 KT IAS, Mach 0.83510 KT TAS, 24000 ft

225 KT IAS, Mach 0.61360 KT TAS, 45º nose upMAX UPWARD VELOCITY

140 KT IAS, Mach 0.43245 KT TAS, 34000 ftBELOW UNACCELERATED STALL SPEED

35

Rotational dynamics The aircraft is rotating through 90º every 30 seconds

Our paper claims these are “barely at the threshold of detection” of the semicircular canals!

Angular velocity of 3 º/s

Centripetal acceleration of 0.006 g

Angular acceleration of 2 º/s2 during transitions between 0 g and 1.8 g

Tangential acceleration of 0.07 g

Karmali F, Shelhamer M. The dynamics of parabolic flight: flight characteristics and passenger percepts. Acta Astronautica (In press)

36

Otolith stimulation

37

Why they call it the “Vomit Comet” Approximately 40-60% of participants get sick on their first

flight. The next day, most are fine.

A sensorimotor rearrangement has occurred. This learning is retained for months.

We want to improve our understanding of these adaptive processes.

Studied a range of sensorimotor responses at different neural levels: brainstem through to perceptual • based on design MVL Spacelab experiments

38

Background

39

Vestibular-Ocular Reflexes

Otolith organs measure linear acceleration and gravity (contains utricle and saccule)

Semicircular canals (SCC) measure angular velocity

Eye movements opposite of head movements

Conjugate: eyes move together

Disconjugate: difference between left and right eyes

VOR gain=Eye velocity / head velocity

40

Eye misalignments occur in parabolic flight and whole-body pitch rotation – otolith implicated (Markham, Diamond, Karmali)

Ocular counterroll is reduced after space flight (Clarke 2000,

Moore 2003, Vogel 1986).

Pitch VOR still exists when canals deactivated (Angelaki)

Otolith-dependent misalignments can be adapted with static head positioning using a visual-vestibular mismatch (Schor)

Eye movements and otolith organs

41

Hypotheses1. Context-specific adaptation: learning a set of

specific responses that is calibrated for each g level.• Example: initially inappropriate otolith-driven responses, which

eventually become appropriate (correctly calibrated) in each g level.

2. Generalized adaptation: learning a set of general responses, each of which is simultaneously appropriate for multiple g levels.• Example: gradual decrease in otolith-driven responses, reflecting

overall lack of reliability of otolith information as g level varies.

42

Methods

43

Experimental design Three consecutive days of flying per subject Responses tested:

• Pre-parabolas• Early• Late• Post-parabolas

14 subjects: 5 experienced, 6 naïve, 3 pilots Measurements:

• torsional alignment • ocular counterroll• vertical alignment • pitch VOR gain – active & passive• Also tried linear VOR, SVV, SPV

No medication for motion sickness

44

Experimental designDay 1 Day 2 Day 3

Pre-parabolas Pre-parabolas Pre-parabolas

Early (first 10 parabolas)



Late (last 10 parabolas)



Post-parabolas Post-parabolas Post-parabolas

45

Measuring pitch VOR gain Active sine-like head

movements • 0.3-1.6 Hz. Results

based on 0.6-1.3 Hz• 20-30º• 40-90º/sec

Fixating stationary point in the light.

Eye movements were recorded using a head-mounted video system (Chronos).

Gain computed by least-squares fitting of eye velocity to head velocity, for each cycle of motion.

46

Measuring static eye position• OCR• Torsional alignment• Vertical alignment

• Nikon D70 digital camera

• Subjects’ head upright/ tilted

47

Results Torsional alignment instability Ocular counterroll (response to head tilt) Active pitch VOR

48Experienced Day 1 Day 2 Day 3

0

0.5

1

1.5

2

2.5In

stab

ility

of to

rsio

nal a

lignm

ent (

deg)


Pilot Subjects(n=3)




Pilot Subjects(n=3)




Pilot Subjects(n=3)



49None Single episode vomiting Severe vomiting

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Motion sickness score

Inst

abilit

y of

tors

iona

l alig

nmen

t (de

g)


500 g 1 g 1.8 g

-8

-6

-4

-2

0

2

g level

Con

juga

te to

rsio

nal e

ye p

ositi

on(+

CW

; deg

rees

)


51

10 15 20 25 30 35 40 45 50 55 60-20

-15

-10

-5

0

5

10

15

20

25

10 15 20 25 30 35 40 45 50 55 60

0.6

0.8

1

1.2

10 15 20 25 30 35 40 45 50 55 600

0.5

1

1.5

2

Eye

& h

ead

posi

tion

VOR G

ain

g le

vel 1.8 g 0 g

520 g 1 g 1.8 g

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)


Day 1 Early

530 g 1 g 1.8 g

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)

Naive subjects (n=3)Pilot subjects (n=2)Day 1

Late

54

Day 3

0 g 1 g 1.8 g0.5

0.6

0.7

0.8

0.9

1

1.1

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)

Naive subjects (n=3)Pilot subjects (n=2)Day 3

55

Conclusions All responses showed a g-level dependence early in flight,

which decreased with experience.

Rate of adaptation varied between reflexes: torsional disconjugacy is fastest, then pitch VOR, then ocular counterroll.

• Torsional alignment instability rapidly reduced upon exposure to parabolic flight, and adaptation is retained between flights. The relatively rapid adaptation may be because errors in torsional alignment have greater functional offsets than conjugate changes in torsional eye position and changes in the pitch VOR.

• Pitch VOR gain initially dropped in 0 g and increased in 1.8 g, consistent with an otolith contribution. Difference between g levels decreased with experience and eventually disappeared, showing that the different otolith contributions in the different g levels are correctly processed after adaptation.

• Ocular counterroll is initially larger in 1.8 g and smaller in 0 g. Differences between g levels do not change within 3 days, although pilots show more appropriate responses in 1 g. An explanation for this slow adaptation may be that a well-tuned OCR gain is not critical; it is not compensatory in normal circumstances.

• None of the mechanisms show a change in response to 1 g after exposure to parabolic flight. This suggests that adaptation is context specific.

Upon adaptation, torsion, torsional alignment and pitch VOR are correctly calibrated in each g level, supporting the hypothesis of a context-specific adaptation of each response.

57

Acknowledgements

Advisor: Dr. Mark Shelhamer Undergraduate researchers: Ondrej Juhasz, Michelle

Zwernemann, Anton Aboukhalil Technical assistance: Adrian Lasker, Dale Roberts,

Dr. Andy Clarke NASA staff: Noel Skinner, John Yaniec Scientific advise: Dr. Mark Walker, Dr. Howard Ying,

SPH Biostats Clinic Moral support: Zee lab, friends & family Funding: NSBRI, NIH, NSERC

58

Compare t-test significance of day 1 vs day 3 naïve

Change bar graphs to std err instead of std

Review adapt. Spaceflight See MJS PPT Inc cool videos ~45 slides

63

Otolith ambiguity – same sensor transduces tilt and translation of head Transduced signal during tilt differs dramatically in weightlessness / high g environments, and changes in

otolith-dependent reflexes occur Designing a countermeasure to speed adaptation or pre-adapt reflexes could reduce adverse problems during

space missions, return to Earth or visiting other celestial bodies Characterizing the adaptive characteristics of otolith-dependent reflexes and the relationship between

adaptation of translation and tilt important for the design of countermeasures To learn more about these pathways, we studied the vestibulo-ocular reflex Specifically interested in the pitch VOR, because this is the most common type of head movement in which

the otolith organs transduce a changing direction of gravity During pitch head movements, three (or more) sensory signals available: otolith-transduced head position;

SCC-transduced head velocity; eye position Goals:

• to adapt VOR and show that adaptation is otolith-organ dependent• Show that adaptation of VOR transfers from tilt to translation• Characterize how the brain processes otolith organ information using the characteristics of the transfer of

adaptation With pitch head movements, during translation and tilt, the compensatory eye movement is vertical. Makes it

impossible to distinguish adaptation that is dependent on eye- vs. head- motion-dependent adaptation. Using vertical eye misalignment allows dependency on velocity and position to be better discriminated.

Methods• How to adapt VOR / eye misalignment?

Results / Discussion Conclusions

• The evidence that Cartesian components of the g vector are used as adaptive cues is important because it suggests that countermeasures on Earth can adapt some responses without changing the magnitude of the g vector. That means adaptation of otolith-dependent reflexes could occur in a 1 g field, which allows longer, cheaper adaptation sessions than jet or parabolic flight.

64

Transfer of oculomotor adaptation between otolith-dependent tilt and translation reflexes

Faisal KarmaliPost-doctoral fellowJVPL / MEEI / HMS

MVL, MITNovember 14, 2007

65

The bottom line…1. A vertical misalignment between the eyes can be

adapted that is dependent on otolith-transduced head tilt during pitch head rotation using a visual-vestibular mismatch

2. Modification of the response to otolith-transduced head tilt also modifies the response to vertical translation

3. Modeling shows that for adaptation, the brain processes the g vector as Cartesian components, rather than in polar coordinates

66

Background

67







VOR gain=Eye velocity / head velocity

68

Eye misalignments occur in parabolic flight and whole-body pitch rotation – otolith implicated (Markham, Diamond, Karmali)

Amount of misalignment decreases with experience in parabolic flight – shows adaptation of pathway occurs

Pitch VOR gain changes with g level – otolith implicated

Pitch VOR still exists when canals deactivated (Angelaki)

Otolith-dependent misalignments can be adapted with static head positioning using a visual-vestibular mismatch (Schor)

Eye movements and otolith organs

69

Previous studies of transfer between translation and rotation Yaw adaptation affects interaural translation

response (Koizuka et al.)

Interaural adaptation affects yaw response (Koizuka et al.)

Yaw adaptation affects response to constant-velocity pitch rotation (Petropoulos, Wall III, Oman)

These studies suggest adaptation of a pathway common to the SCC and otolith reflexes

In contrast, we find adaptation specific to the otolith pathway

70

0 g 1 g 1.8 g0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

g level

Gai

n of

pitc

h ve

stib

uloo

cula

r ref

lex

(deg

/deg

; ide

al g

ain

is 1

.0)


0 g 1 g 1.8 gg level

Naive subjects (n=3)Pilot subjects (n=2)

Day 1, early Day 1, late

Pitch VOR gain is dependent on g level in parabolic flight

71

Experimental Design

72

Aims1. Show that a misalignment between the eyes can be

adapted that it dependent on otolith-transduced tilt during dynamic pitch rotation using a visual-vestibular mismatch

2. Determine if modification of the response to otolith-transduced head tilt also modifies the response to vertical translation

3. Develop a model to understand how the brain processes otolith information during adaptation of otolith-ocular responses

73

Motion profiles

0 20 40 60 80 100 120-4

-2

0

2

4

Act

ive

pitc

h ro

tatio

n

0 20 40 60 80 100 120-4

-2

0

2

4

Ver

tical

sle

d tra

nsla

tion

Time (seconds)

74

Motion profileSensory receptor

Active pitch rotation Vertical translation

Otolith organs

Changing orientation relative to gravity

Linear acceleration

Semicircular canals (SCC)

Yes No

Orbital eye position

Yes Yes

Collic (neck) reflex

Yes No

Sensory environment during motion profiles

75


Passive pitch rotation (full body)


Otolith organs Changing orientation relative to gravity


Linear acceleration


Only before vestibular time constant exceeded

Yes No


Yes, due to counterpitch response

Yes Yes


No Yes No

Cognitive intent No Yes NoCognitive anticipation

Some situations Yes Yes


76

Pre-test Pitch(monocular targets)

Adapt Pitch(20 min)

Post-test Pitch (monocular targets)

Pre-test Vertical (monocular target)

Pre-test Pitch (monocular target)

Adapt Pitch(15 min)

Post-test Pitch (monocular target)

Adapt Pitch(5 min)

Post-test vertical(monocular target)

The pitch plasticity study

The vertical translate study

Study procedures

77

Scleral contact lens coils Coil frame in a cube that consists of

three orthogonal magnetic fields oscillating at different frequencies

Coil acts as antenna that picks up a linear combination of the three that depends on its orientation

Accuracy <0.01º Wire is fragile, and often breaks Small coil frame used on vertical sled – data for pitch

not usable during the vertical translate study

78

Generalized Estimating Equations

t-test is often used to compare time-series data in two different conditions

t-test assumes independence between measures, which is not usually the case with rapidly-changing, sampled data such as eye position

GEE is a more correct and stringent technique that takes into account the correlation of the data with itself

t-test will sometimes indicate significance when GEE does not

Zeger, S. L. & Liang, K. Y. (1986). Longitudinal data analysis for discrete and continuous outcomes. Biometrics 42, 121-130.Karmali, Ramat, Shelhamer (2006). Journal of Vestibular Research 16:117-125.

79

Goal: understand the nature of central compensation Study the ability of the brain to learn an otolith organ-dependent

misalignment during dynamic head movements Subject performs active pitch rotation Induced a vertical misalignment of the eye by presenting a

visual disparity between the left and right eyes Subject wears red-blue glasses Real-time head position is fed to laptop which projects a field of

dots onto a screen The dots have a vertical disparity between left and right eyes Regular & eye-head dissociation paradigms

Pitch plasticity study

80


95 100 105 110 115 120

-20

0

20

40

Hea

d po

sitio

n(d

egre

es; +

up)

95 100 105 110 115 120

-40

-20

0

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

up) Right eye

Left eye

95 100 105 110 115 120-1

0

1

2

3

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

95 100 105 110 115 120

-0.1-0.05

00.050.1

Mis

alig

nmen

t gai

n

Time (seconds)

1645 1650 1655 1660 1665 1670

-20

0

20

40

Hea

d po

sitio

n(d

egre

es; +

up)

1645 1650 1655 1660 1665 1670

-40

-20

0

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

up)

1645 1650 1655 1660 1665 1670-1

0

1

2

3

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

1645 1650 1655 1660 1665 1670

-0.1-0.05

00.050.1

Mis

alig

nmen

t gai

n

Time (seconds)

Pre-adaptation Post-adaptation

81-50 0 50

-4

-3

-2

-1

0

1

2

3

4

Misalignment vs. head position forthree target positions (subject J)

Verti

cal m

isal

ignm

ent (

degr

ees;

righ

t-lef

t; +

up)

Head position (degrees; + up)

Pre-adaptation

Post-adaptation

82

-50 0 50-4

-2

0

2

4Subject HSubject HSubject HSubject HSubject HSubject H

Standard paradigm

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject H*Subject H*Subject H*Subject H*Subject H*Subject H*

Eye-head dissociation paradigm

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject JSubject JSubject JSubject JSubject JSubject J

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject J*Subject J*Subject J*Subject J*Subject J*Subject J*

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject LSubject LSubject LSubject LSubject LSubject L

Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject M*Subject M*Subject M*Subject M*Subject M*Subject M*

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2

4Subject BSubject BSubject BSubject BSubject BSubject B

Head position(degrees; + up)

-50 0 50-4

-2

0

2


Eye position(deg; + down)

-50 0 50-4

-2

0

2

4Subject B*Subject B*Subject B*Subject B*Subject B*Subject B*


-50 0 50-4

-2

0

2



-50 0 50-4

-2

0

2

4Subject YSubject YSubject YSubject YSubject YSubject Y

-50 0 50-4

-2

0

2


83

Position-dependent misalignment increased in most experiments

* eye-head dissociation paradigm

H H* J J* L M* B B* Y-0.01

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

**

**

**

**

**

**

**

**

Coe

ffici

ent o

f cha

nge

in p

ositi

on-d

epen

dent

mis

alig

nmen

t fro

m p

re to

pos

t (de

g/de

g)

Subject

0

10

20

30

40

50

60

% o

f req

uest

ed g

ain

chan

ge

** p<0.01; GEE

84-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Head position (deg; + up)

Ideal results demonstrating eye position-dependent adaptation

-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Eye position (deg; + down)

-50 0 50-4

-3

-2

-1

0

1

2

3

4V

ertic

al m

isal

ignm

ent

(deg

rees

; rig

ht-le

ft; +

up)


Ideal results demonstrating head position-dependent adaptation

-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)


Up target

fixatio

n

Center ta

rget fi

xatio

n

Down targ

et fixati

on

85

Example: misalignment dependent on head position

-50 0 50-4

-3

-2

-1

0

1

2

3

4

Misalignment vs. head position forthree target positions (subject H)

Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)

Head position (degrees; + up)-50 0 50

-4

-3

-2

-1

0

1

2

3

4

Misalignment vs. eye position forthree target positions (subject H)

Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)

Eye position (degrees; + down)

index of head-position-dependent adaptationheadeye

headeye

MSMSMSMS

+−

=

86

Example: misalignment partly dependent on eye position

-50 0 50-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


87

Is adaptation dependent on eye or head? Compare mean sum-of-squares of regression for

misalignment on eye, and misalignment on head

H H* J J* L M* B B* Y-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

Inde

x of

hea

d-po

sitio

n-de

pend

ent a

dapt

atio

n

Subject

Subject exhibits head-position-dependent adaptation

Subject exhibits eye-position-dependent adaptation

88

Pitch plasticity study – summary Position-dependent misalignments were adapted in

all subjects

Head position was implicated as the adaptation cue in most subjects

Otolith organs or SCC could be driving adaptation

In the next study, we performed the same adaptation and also tested in vertical translation, which stimulated the otolith organs but not the SCC

89

Vertical translate studyPre-test Vertical (monocular target)


Adapt Pitch(15 min)


Adapt Pitch(5 min)

Post-test vertical(monocular target)

900.95 Hertz ±0.7 g peak acceleration

4 6 8 10 12 14 16 18-10

-5

0

5

10

Upward

Downward

Time (seconds)Ver

tical

acc

eler

atio

n (m

/s2 ) Vertical sled movement characteristics

4 6 8 10 12 14 16 18-2

-1

0

1

2

Time (seconds)

Ver

tical

vel

ocity

(m/s

)

4 6 8 10 12 14 16 18

-0.2

-0.1

0

0.1

0.2

0.3

Top

Bottom

Time (seconds)

Ver

tical

pos

ition

(m)

91

Response to vertical translation

170 175 180 185 190

-5

0

5

Hea

d ac

cele

ratio

n(m

/s/s

; + u

p)

Pre-adaptation response (subject J)

170 175 180 185 1900

2

4

6

8

Ver

genc

e an

gle

(deg

rees

)

170 175 180 185 190-20

-10

0

10

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

dow

n)

Right eyeLeft eye

170 175 180 185 190-2

-1

0

1

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Time (seconds)

2050 2055 2060 2065 2070

-5

0

5

Hea

d ac

cele

ratio

n(m

/s/s

; + u

p)

Post-adaptation response (subject J)

2050 2055 2060 2065 20700

2

4

6

8

Ver

genc

e an

gle

(deg

rees

)2050 2055 2060 2065 2070

-20

-10

0

10

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

dow

n)

2050 2055 2060 2065 2070-2

-1

0

1

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Time (seconds)

92

Sensitivity to vertical translation increases after adaptation of pitch response

Increase in sensitivity increases for 4 of 5 subjects

Change statistically significant in all subjects (GEE)

-1 -0.5 0 0.5 1-1

0

1Sensitivity 0.083°/g (subject H)

-1 -0.5 0 0.5 1-1

0

1Sensitivity change 0.084°/g

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity 0.190°/g (subject J)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity 0.289°/g (subject M)

Verti

cal m

isal

ignm

ent (

degr

ees;

righ

t-lef

t; +

up)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity -0.053°/g (subject B)

-1 -0.5 0 0.5 1-1

0

1Sensitivity change -0.078°/g

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity -0.217°/g (subject W)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity 0.053°/g (subject W)

Vertical acceleration (g; 1 g=9.81 m/s2)

Pre-adaptation Post-adaptation Difference

93


-0.05 0 0.05 0.1-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Gain change for active pitch rotation (°/°)Sens

itivi

ty c

hang

e fo

r ver

tical

tran

slat

ion

(°/g

)

H

J

B

M

94

Vertical translate study – summary

Head-position-dependent adaptation of misalignment during pitch rotation was otolith organ dependent

Adaptation transferred from otolith-dependent tilt response to otolith-dependent translation response

Misalignment is a tool to study the otolith-ocular pathway: it allows questions to be asked about tilt/translation which are difficult to answer using conjugate eye movements

95

What model of otolith organ information processing explains the transfer of

adaptation from pitch to vertical translation?

pitchangle

g vector

g vector

VT

NO

Polar coordinates

Cartesian components

VT

g vector

NO

96

0 20 40 60 80 100-20

0

20Pitch Plasticity Study

Hea

d pi

tch

angl

e(d

egre

es)

0 20 40 60 80 100-2

0

2

Car

tesi

anco

mpo

nent

s(g

= 9

.81

m/s

/s)

NOVT

0 20 40 60 80 100-4-2024 misalignment=c*pitchangle [c=-1/10]

Pol

ar

0 20 40 60 80 100-4-2024 misalignment=a*NO+b*VT [a=-4; b=0]

Mis

alig

nmen

t pre

dict

ed b

yC

arte

sian

NO

onl

y


Car

tesi

anN

O a

nd V

T(n

on-o

ptim

al)

0 20 40 60 80 100-4-2024 misalignment=b*(k*NO+VT) [b=1; k=-4]

Car

tesi

an(s

ingl

e kn

ob)

Time (seconds)

0 20 40 60 80 100-1

0

1Vertical Translate Study

0 20 40 60 80 100-2

0

2

0 20 40 60 80 100-4-2024 misalignment=c*pitchangle [c=-1/10]



0 20 40 60 80 100-4-2024 misalignment=b*(k*NO+VT) [b=1; k=4]

Time (seconds)

97

Modeling – summary

Otolith organ information is processed in Cartesian components

Adaptation affects all g vector components equally

98

Conclusions Otolith-dependent eye misalignments can be

adaptively created during dynamic pitch rotation using a visual-vestibular mismatch

Adaptation transfers from otolith-transduced tilt responses to translation responses

Components of the g vector are used to determine the ocular response; adaptation affects all components equally

99

Application to countermeasures Adapting otolith reflexes during weightlessness is

expensive and of a short duration

Countermeasures can focus on adapting responses of individual Cartesian components

When appropriate g vector cannot be provided using translation, it can be provided using tilt

100

Future work Test for transfer when adapted Cartesian

components are the same

The relationship between head position and misalignment in the adaptive paradigm was linear –a parabolic relationship would separate otolith and SCC contributions

Transfer of adaptation from vertical translation to pitch rotation

101

Future work

Head orientation during adaptation during first test during second test rotation axis g vector

components stimulated

rotation axis g vector components stimulated


Pitch upright VT & NO Roll onside VT (&IA) Yaw onside NO (&IA) Roll upright VT & IA Pitch supine VT (&NO) Yaw supine IA (&NO) Yaw supine NO & IA Pitch upright NO (&VT) Roll upright IA (&VT)

102

Acknowledgements

Advisor: Dr. Mark Shelhamer Committee members: Drs. Paul

Fuchs, David Solomon, David Zee, Kechen Zhang

Technical assistance: Adrian Lasker, Dale Roberts, NASA staff, Dr. Andy Clarke

Scientific advise: Dr. Mark Walker, Dr. Howard Ying, School of Public Health Biostats Clinic

Funding: NSBRI, NIH, NSERC

105

Vertical eye misalignments during pitch rotation and vertical translation:

Evidence for bilateral asymmetries and plasticity in the otolith-ocular reflex

Dr. Faisal Karmali MEEI

September 10, 2007

106

Publications 12 conference abstracts Peer-reviewed publications

• Karmali F, Ramat S, Shelhamer M. Vertical skew due to changes in gravitoinertial force: A possible consequence of otolith asymmetry. Journal of Vestibular Research. 2006 Dec;16:117-125.

• Karmali F, Shelhamer M. Automatic Detection of Camera Translation in Eye Video Recordings using Multiple Methods. Annals of the New York Academy of Sciences. 2005 Apr;1039:470-6.

• Karmali F, Shelhamer M. Compensating for camera translation in video eye movement recordings by tracking a landmark selected automatically by a genetic algorithm. (submitted)

• Karmali F, Shelhamer M. The dynamics of parabolic flight: flight characteristics and passenger percepts. (submitted)

• Grabherr L, Karmali F, Bach S, Indermaur K, Metzler S, Mast FW. Mental Rotation of Bodies and Body-parts in Microgravity. Journal of Vestibular Research. 2007. (accepted)

• Karmali F, Ramat S, Straumann D, Shelhamer M. Binocular disconjugacy in slow and fast pitch vestibulo-ocular reflex: Evidence for otolithic influence. (in preparation)

107

Background

108

The Vestibular System “Sixth sense” that keeps us balanced Stops us from falling when we stumble Helps move eyes (“gaze stabilization”)

Patients with Vestibular Disease or stroke

Astronauts affected by weightlessness

VertigoDifficulty walkingEyes move incorrectly

Vertigo and motion sicknessDifficulty walking upon returnEyes move incorrectly

http://accuweather.ap.org/cgi-bin/apdownload.pl?6754643+Intl_Photos+accuweather.ap.org:80+++�

109





Torsion is rotation of eye about the line of sight



110

The otolith organ: a mass on a lever

The lever is a hair cell that measures deflection

The mass is called the otoconia and deflects when acted upon by external forces

g vector: the sum of linear acceleration and gravity 1 g of downward gravity is indistinguishable from 1 g of upward acceleration of the head

Temporal bone of the

head

111

Vertical misalignments during parabolic flight

Karmali, Ramat, Shelhamer, Journal of Vestibular Research. 2006 Dec;16:117-125

10 20 30 40 50 60

-5

0

5

Left Eye

Right Eye

Ver

tical

eye

posi

tions

(deg

rees

) y g g Right EyeLeft Eye

10 20 30 40 50 60-10

-5

0

5

10

Com

pens

ated

ver

tical

eye

po

sitio

ns (d

egre

es)

Right EyeLeft Eye

10 20 30 40 50 60

0

1

2

Time (seconds) Trial: D1

Gra

vito

iner

tial

acce

lera

tion

(g)

112

Otolith Asymmetry Hypothesis

Left Otolith

Right Otolith

Left Eye

Right Eye

g vector

Central Compensation

113

“Over 50% of utricular-activated, second-order vestibular neurons received commissural inhibition from the contralateral utricular nerve.”

“Almost all the saccular-activated, second-order vestibular neurons exhibit no response to stimulation of the contralateral saccular nerve.”

Uchino, 2004

114

Implications of otolith asymmetry1. It can predict space sickness: a better understanding

of the mechanisms may help to improve screening and produce simpler screening tests

2. It may reduce task performance by creating a sensory conflict or misaligning the eyes

3. Understanding how it adapts may be useful in training including partial adaptation before flight

115

Experimental Design

116

Motion profiles

0 20 40 60 80 100 120-4

-2

0

2

4

Par

abol

ic fl

ight Pitch angular position

g vector

0 20 40 60 80 100 120-4

-2

0

2

4

Pas

sive

pitc

h ro

tatio

n

0 20 40 60 80 100 120-4

-2

0

2

4

Act

ive

pitc

h ro

tatio

n

0 20 40 60 80 100 120-4

-2

0

2

4

Ver

tical

sle

d tra

nsla

tion

Time (seconds)

117






Linear acceleration



Yes No



Yes Yes


No Yes No


118






Linear acceleration



Yes No



Yes Yes


No Yes No

Cognitive intent No Yes NoCognitive anticipation

Some situations Yes Yes


119

Slow test Pitch

Medium testPitch

Fast test Pitch

Pre-test Pitch (monocular targets)

Adapt Pitch (20 min)

Post-test Pitch (monocular targets)





Adapt Pitch (5 min)

Post-test vertical (monocular target)

The pitch innate study

The pitch plasticity study


Study procedures

120

Scleral contact lens coils Coil frame in a cube that consists of

three orthogonal magnetic fields oscillating at different frequencies

Coil acts as antenna that picks up a linear combination of the three that depends on its orientation

Accuracy <0.01º Wire is fragile, and often breaks Small coil frame used on vertical sled – data for pitch

not usable during the vertical translate study

121

Generalized Estimating Equations

t-test is often used to compare time-series data in two different conditions

t-test assumes independence between measures, which is not usually the case with rapidly-changing, sampled data such as eye position

GEE is a more correct and stringent technique that takes into account the correlation of the data with itself

t-test will sometimes indicate significance when GEE does not

Zeger, S. L. & Liang, K. Y. (1986). Longitudinal data analysis for discrete and continuous outcomes. Biometrics 42, 121-130.Karmali, Ramat, Shelhamer, Journal of Vestibular Research. 2006 Dec;16:117-125

122

Pitch innate study

coils

123

Static alignment

0 0.5 1-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Time (seconds)

Subject AV

ertic

al m

isal

ignm

ent (

Deg

rees

, rig

ht-le

ft, +

dow

n)

0° (τ=3.26 s)least-squares fit270° (τ=0.35 s)least-squares fit180° (τ=0.00 s)least-squares fit90° (τ=0.32 s)least-squares fit

0 0.5 1-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Time (seconds)

Subject T


0 0.5 1-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Time (seconds)

Subject S


0 0.5 1-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Time (seconds)

Subject E


0 0.5 1-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Time (seconds)

Subject Q


124

0 20 40 60 80 100 120 140-2

0

2

4

6

8

Ver

tical

eye

pos

ition

s (d

egre

es, +

dow

n)

Vertical misalignment during slow pitch rotation (6 deg/sec)

Right eyeLeft eye

0 20 40 60 80 100 120 1400

200

400BackwardRotation

ForwardRotation

Cha

ir po

sitio

n (d

egre

es)

-slo

pe p

itch

forw

ard

Time

Slow rotation – 6º/sec

125

Medium rotation – 60º/sec

0 5 10 15 20 25 30 35 40 45 50 55-20

-10

0

10

20

Time (seconds)

Ver

tical

eye

pos

ition

s(d

egre

es, +

dow

n) Right eyeLeft eyeChair positionTarget flash

0 5 10 15 20 25 30 35 40 45 50 55-1

-0.5

0

0.5

1

Time (seconds)

Ver

tical

mis

alig

nmen

t(d

egre

es, r

ight

-left,

+do

wn)

Vertical misalignmentChair positionTarget flash

0 50 100 150 200 250 300 350-1.5

-1

-0.5

0

0.5

1

1.5

Chair position (degrees)

Ver

tical

mis

alig

nmen

t(d

egre

es, r

ight

-left,

+do

wn)

Upright

Upside-down

Upright

126

0 1000 2000 3000 4000 5000 6000 7000 8000-100

-80

-60

-40

-20

0

Ver

tical

eye

pos

ition

s (d

egre

es, +

dow

n)

Right eyeLeft eyeChair position (inverted)

0 1000 2000 3000 4000 5000 6000 7000 8000-2

-1.5

-1

-0.5

0

0.5

1

Fixa

tion

targ

et o

ff

Fixa

tion

targ

et o

n

Ver

tical

mis

alig

nmen

t(d

egre

es, r

ight

-left,

+do

wn)

0 1000 2000 3000 4000 5000 6000 7000 8000-100

-50

0

50

100

Ver

tical

eye

vel

ocity

(deg

/s, +

dow

n)

0 1000 2000 3000 4000 5000 6000 7000 8000-4

-2

0

2

4

Time (ms)

Mis

alig

nmen

t vel

ocity

(deg

/s, r

ight

-left,

+do

wn)

Fast rotation – 2 second steps

127

-3000 -2000 -1000 0 1000 2000 3000 4000 5000 6000-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

Time (msec)

Ver

tical

mis

alig

nmen

t(d

egre

es, r

ight

-left,

+do

wn)

Forward movement

Fixa

tion

targ

et o

ff

Mov

emen

tco

mm

ence

s

Fixa

tion

targ

et o

n

-3000 -2000 -1000 0 1000 2000 3000 4000 5000 6000-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

Time (msec)

Ver

tical

mis

alig

nmen

t(d

egre

es, r

ight

-left,

+do

wn)

Backward movement

Fixa

tion

targ

et o

ff

Mov

emen

tco

mm

ence

s

Fixa

tion

targ

et o

n

upright→supinesupine→upside-downupside-down→proneprone→upright


128

Backward movement

uprig

ht→

supi

ne

supi

ne→

upsi

de-d

own

upsi

de-d

own→

pron

e

pron

e→up

right

ATSEQ

-5

-4

-3

-2

-1

0

1

2

3

4

5

g

g(d

egre

es, r

ight

-left,

+do

wn)

Forward movement

uprig

ht→

pron

e

pron

e→up

side

-dow

n

upsi

de-d

own→

supi

ne

supi

ne→

uprig

htATSEQ


129

A E Q R S T

-3

-2

-1

0

1

2

3

**

**

**

*

**

**

ns

**

ns

Fast

: pea

k m

isal

ignm

ent (

forw

ard)

Fast

: pea

k m

isal

ignm

ent (

back

war

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(fo

rwar

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(ba

ckw

ard)

Sta

tic o

rient

atio

n: m

isal

ignm

ent d

iffer

entia

lS

low

: pea

k ve

rtica

l mis

alig

nmen

t (fo

rwar

d)S

low

: pea

k m

isal

ignm

ent (

back

war

d)M

ediu

m: m

isal

ignm

ent d

iffer

entia

l (fo

rwar

d)M

ediu

m: m

isal

ignm

ent d

iffer

entia

l (ba

ckw

ard)

**

**

**

*

**

*

**

Fast

: pea

k m

isal

ignm

ent (

forw

ard)

Fast

: pea

k m

isal

ignm

ent (

back

war

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(fo

rwar

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(ba

ckw

ard)

Sta

tic o

rient

atio

n: m

isal

ignm

ent d

iffer

entia

lS

low

: pea

k ve

rtica

l mis

alig

nmen

t (fo

rwar

d)S

low

: pea

k m

isal

ignm

ent (

back

war

d)

** **

*

ns

ns

**

ns

Fast

: pea

k m

isal

ignm

ent (

forw

ard)

Fast

: pea

k m

isal

ignm

ent (

back

war

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(fo

rwar

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(ba

ckw

ard)

Sta

tic o

rient

atio

n: m

isal

ignm

ent d

iffer

entia

lS

low

: pea

k ve

rtica

l mis

alig

nmen

t (fo

rwar

d)S

low

: pea

k m

isal

ignm

ent (

back

war

d)M

ediu

m: m

isal

ignm

ent d

iffer

entia

l (fo

rwar

d)

**

Med

ium

: mis

alig

nmen

t diff

eren

tial (

forw

ard)

Med

ium

: mis

alig

nmen

t diff

eren

tial (

back

war

d)

**

ns

ns

ns

**

Fast

: pea

k m

isal

ignm

ent (

forw

ard)

Fast

: pea

k m

isal

ignm

ent (

back

war

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(fo

rwar

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(ba

ckw

ard)

Sta

tic o

rient

atio

n: m

isal

ignm

ent d

iffer

entia

l

ns

**

ns

**

Fast

: pea

k m

isal

ignm

ent (

forw

ard)

Fast

: pea

k m

isal

ignm

ent (

back

war

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(fo

rwar

d)Fa

st: p

eak

mis

alig

nmen

t vel

ocity

(ba

ckw

ard)

Sta

tic o

rient

atio

n: m

isal

ignm

ent d

iffer

entia

l

Med

ium

: mis

alig

nmen

t diff

eren

tial (

back

war

d)

Subject

Verti

cal m

isal

ignm

ent (

Deg

rees

, rig

ht-le

ft, +

dow

n)Summary of vertical eye misalignments for static orientation and slow, medium and fast pitch rotations

Static orientation: misalignment differentialSlow: peak vertical misalignment (forward)Slow: peak misalignment (backward)Medium: misalignment differential (forward)Medium: misalignment differential (backward)Fast: significance of orientation dependence - peak misalignment (forward)Fast: significance of orientation dependence - peak misalignment (backward)Fast: significance of orientation dependence - peak misalignment velocity (forward)Fast: significance of orientation dependence - peak misalignment velocity (backward)

130

Pitch innate study - summaryMisalignment determined by

Average misalignment difference between upright & upside-down

Number of significant subjects (p<0.01; GEE)

Static Orientation 0.99º 3/5

Slow Orientation 1.74º 3/3

Medium Orientation 0.39º 3/4

Fast Angular velocity (orientation significant in 1/5 subjects)

1.86º (peak misalignment) Velocity-dependent

misalignments implicate SCC pathway

Orientation-dependent misalignments implicate asymmetry in otolith pathway

Otolith asymmetry acts in both:• pitch rotation - g vector direction changing• vertical acceleration - g vector magnitude changing

131

Goal: understand the nature of central compensation Study the ability of the brain to learn an otolith organ-dependent

misalignment during dynamic head movements Subject performs active pitch rotation Induced a vertical misalignment of the eye by presenting a

visual disparity between the left and right eyes Subject wears red-blue glasses Real-time head position is fed to laptop which projects a field of

dots onto a screen The dots have a vertical disparity between left and right eyes Regular & eye-head dissociation paradigms


132


95 100 105 110 115 120

-20

0

20

40

Hea

d po

sitio

n(d

egre

es; +

up)

95 100 105 110 115 120

-40

-20

0

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

up) Right eye

Left eye

95 100 105 110 115 120-1

0

1

2

3

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

95 100 105 110 115 120

-0.1-0.05

00.050.1

Mis

alig

nmen

t gai

n

Time (seconds)

1645 1650 1655 1660 1665 1670

-20

0

20

40

Hea

d po

sitio

n(d

egre

es; +

up)

1645 1650 1655 1660 1665 1670

-40

-20

0

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

up)

1645 1650 1655 1660 1665 1670-1

0

1

2

3

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

1645 1650 1655 1660 1665 1670

-0.1-0.05

00.050.1

Mis

alig

nmen

t gai

n

Time (seconds)

Pre-adaptation Post-adaptation

133-50 0 50

-4

-3

-2

-1

0

1

2

3

4

Misalignment vs. head position forthree target positions (subject J)

Verti

cal m

isal

ignm

ent (

degr

ees;

righ

t-lef

t; +

up)

Head position (degrees; + up)

Pre-adaptation

Post-adaptation

134

-50 0 50-4

-2

0

2


Standard paradigm

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


Eye-head dissociation paradigm

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)

-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2



-50 0 50-4

-2

0

2



-50 0 50-4

-2

0

2



-50 0 50-4

-2

0

2



-50 0 50-4

-2

0

2


-50 0 50-4

-2

0

2


135

Position-dependent misalignment increased in most experiments

* eye-head dissociation paradigm

H H* J J* L M* B B* Y-0.01

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

**

**

**

**

**

**

**

**

Coe

ffici

ent o

f cha

nge

in p

ositi

on-d

epen

dent

mis

alig

nmen

t fro

m p

re to

pos

t (de

g/de

g)

Subject

0

10

20

30

40

50

60

% o

f req

uest

ed g

ain

chan

ge

** p<0.01; GEE

136-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)


Ideal results demonstrating eye position-dependent adaptation

-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)


-50 0 50-4

-3

-2

-1

0

1

2

3

4V

ertic

al m

isal

ignm

ent

(deg

rees

; rig

ht-le

ft; +

up)


Ideal results demonstrating head position-dependent adaptation

-50 0 50-4

-3

-2

-1

0

1

2

3

4

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)


Up target

fixatio

n

Center ta

rget fi

xatio

n

Down targ

et fixati

on

137

Example: misalignment dependent on head position

-50 0 50-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


index of head-position-dependent adaptationheadeye

headeye

MSMSMSMS

+−

=

138

Example: misalignment partly dependent on eye position

-50 0 50-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


-4

-3

-2

-1

0

1

2

3

4


Ver

tical

mis

alig

nmen

t (de

gree

s; ri

ght-l

eft;

+ up

)


139

Is adaptation dependent on eye or head? Compare mean sum-of-squares of regression from

misalignment on eye and misalignment on head

H H* J J* L M* B B* Y-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

Inde

x of

hea

d-po

sitio

n-de

pend

ent a

dapt

atio

n

Subject

Subject exhibits head-position-dependent adaptation

Subject exhibits eye-position-dependent adaptation

140

Pitch plasticity study – summary Position-dependent misalignments were adapted in

all subjects Head position was implicated as the adaptation cue

in most subjects Otolith organs or SCC could be driving adaptation In the next study, we performed this experiment and

tested in vertical translation, which stimulated the otolith organs but not the SCC

141

Vertical translate study





Adapt Pitch (5 min)

Post-test vertical (monocular target)


1420.95 Hertz ±0.7 g peak acceleration

4 6 8 10 12 14 16 18-10

-5

0

5

10

Upward

Downward

Time (seconds)Ver

tical

acc

eler

atio

n (m

/s2 ) Vertical sled movement characteristics

4 6 8 10 12 14 16 18-2

-1

0

1

2

Time (seconds)

Ver

tical

vel

ocity

(m/s

)

4 6 8 10 12 14 16 18

-0.2

-0.1

0

0.1

0.2

0.3

Top

Bottom

Time (seconds)

Ver

tical

pos

ition

(m)

143

Response to vertical translation

170 175 180 185 190

-5

0

5

Hea

d ac

cele

ratio

n(m

/s/s

; + u

p)

Pre-adaptation response (subject J)

170 175 180 185 1900

2

4

6

8

Ver

genc

e an

gle

(deg

rees

)

170 175 180 185 190-20

-10

0

10

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

dow

n)

Right eyeLeft eye

170 175 180 185 190-2

-1

0

1

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Time (seconds)

2050 2055 2060 2065 2070

-5

0

5

Hea

d ac

cele

ratio

n(m

/s/s

; + u

p)

Post-adaptation response (subject J)

2050 2055 2060 2065 20700

2

4

6

8

Ver

genc

e an

gle

(deg

rees

)2050 2055 2060 2065 2070

-20

-10

0

10

20

Ver

tical

eye

pos

ition

s(d

egre

es; +

dow

n)

2050 2055 2060 2065 2070-2

-1

0

1

Ver

tical

mis

alig

nmen

t(d

egre

es; r

ight

-left;

+ u

p)

Time (seconds)

144


Increase in sensitivity increases for 4 of 5 subjects

Change statistically significant in all subjects (GEE)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


Verti

cal m

isal

ignm

ent (

degr

ees;

righ

t-lef

t; +

up)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity change -0.078°/g

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity -0.217°/g (subject W)

-1 -0.5 0 0.5 1-1

0


-1 -0.5 0 0.5 1-1

0

1Sensitivity 0.053°/g (subject W)

Vertical acceleration (g; 1 g=9.81 m/s2)

145


-0.05 0 0.05 0.1-0.1

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Gain change for active pitch rotation (°/°)Sens

itivi

ty c

hang

e fo

r ver

tical

tran

slat

ion

(°/g

)

H

J

B

M

146

Vertical translate study – summary

Head-position-dependent adaptation of misalignment during pitch rotation was otolith organ dependent

Adaptation transferred from otolith-dependent tilt response to otolith-dependent translation response

Misalignment is a tool to study the otolith-ocular pathway: it allows questions to be asked about tilt/translation which are difficult to answer using conjugate eye movements

147

Modeling

What is the nature of the central compensation? What model of otolith organ information processing

explains the transfer of adaptation from pitch rotation to vertical translation?

148

Model overview

Right otolith organ

Left otolith organ

Sensorimotor processing

Central compensation

Sensorimotor processing +

+

–

Left eye position

Right eye position

Visual processing of misalignment

Head tilt and translation

Commissure

Leye

Raff

Reye

Lcomm

Rcomm

Laff

–k

–k Vertical eye misalignment

Eye plant

Eye plant

AdditiveMultiplicativeContext-specific adaptation

149

Additive vs.MultiplicativeCompensation -1 0 1 2

Rcomm=(q*g)

Lcomm=g

Pre-commissural neurons

No

com

pens

atio

n

-1 0 1 2

Reye=(q*g)-k*g

Leye=g-k*(q*g)

Post-commissural neurons

-1 0 1 2

Lcomm=g

Rcomm=(q*g-a)

Pre

-com

mis

sura

lad

ditiv

e

-1 0 1 2

Leye=g-k*(q*g-a)

Reye=(q*g-a)-k*g

-1 0 1 2

Raff=q*g

Laff=g

Otolith organ afferents

neur

onal

firin

g ra

te

g level (g)

-1 0 1 2

Lcomm=gRcomm=(q*g*b)

Pre

-com

mis

sura

lm

ultip

licat

ive

-1 0 1 2

Leye=g-k*(q*g*b)Reye=(q*g*b)-k*g

-1 0 1 2

Rcomm=(q*g)

Lcomm=g

Pos

t-co

mm

issu

ral

addi

tive

-1 0 1 2

Leye=g-k*(q*g)-c

Reye=(q*g)-k*g

-1 0 1 2

Rcomm=(q*g)

Lcomm=g

Pos

t-co

mm

issu

ral

mul

tiplic

ativ

e

g level (g)-1 0 1 2

Leye=g-k*(q*g)Reye=d*((q*g)-k*g)

g level (g)

150

What model of otolith organ information processing explains the transfer of

adaptation from pitch to vertical translation?

pitchangle

g vector

g vector

VT

NO

g vector model

g vector components model

VT

g vector

NO

151

Eye torsion with head tilted during change g level Torsion changes even though g vector direction does not

change Suggests the g vector component determines torsion

152

0 20 40 60 80 100-20

0

20Pitch Plasticity Study

Hea

d pi

tch

angl

e(d

egre

es)

0 20 40 60 80 100-2

0

2

g ve

ctor

com

pone

nts

(g =

9.8

1 m

/s/s

)

NOVT

0 20 40 60 80 100-4-2024

misalignment=c*pitchangle [c=-1/10]

g ve

ctor

mod

el

0 20 40 60 80 100-4-2024

misalignment=a*NO+b*VT [a=-4; b=0]

Mis

alig

nmen

t pre

dict

ed b

yg

vect

orco

mpo

nent

s m

odel

0 20 40 60 80 100-4-2024


g ve

ctor

com

pone

nts

mod

el(n

on-o

ptim

al)

0 20 40 60 80 100-4-2024

misalignment=b*(k*NO+VT) [b=1; k=-4]

g ve

ctor

com

pone

nts

mod

el(s

ingl

e kn

ob)

Time (seconds)

0 20 40 60 80 100-1

0

1Vertical Translate Study

0 20 40 60 80 100-2

0

2

0 20 40 60 80 100-4-2024

misalignment=c*pitchangle [c=-1/10]

0 20 40 60 80 100-4-2024


0 20 40 60 80 100-4-2024


0 20 40 60 80 100-4-2024

misalignment=b*(k*NO+VT) [b=1; k=4]

Time (seconds)

153

Modeling – summary

Context-specific adaptation is a likely candidate for central compensation

Otolith organ information is processed in g vector components

Adaptation affects all g vector components equally

154

Conclusions Vertical eye misalignments that are dependent on

the otolith organs occur during pitch rotation Vertical eye misalignments that are dependent on

the SCC occur during fast pitch rotation Otolith-dependent misalignments can be adaptively

created during dynamic pitch rotation using a visual-vestibular mismatch

Context-specific adaptation is a likely candidate for central compensation

Components of the g vector are used to determine the ocular response; adaptation affects all components equally

155

Implications Understanding central compensation in the otolith-

ocular pathway can improve the design of adaptation regiments for both astronauts and patients

Understanding how otolith information is used by the brain can guide the selection of motion profiles used for adaptation paradigms

The ability to measure innate misalignments using pitch rotation may help in assessing risks during spaceflight

156

Future work Model a SCC asymmetry by having canal planes that

are slightly different in the left and right ear The relationship between head position and

misalignment in the adaptive paradigm was linear –a parabolic relationship would help to separate otolith and SCC contributions

In fast rotation, SCC-dependent misalignments were implicated – further investigation is required

Transfer of adaptation from vertical translation to pitch rotation

Transfer of adaptation between otolith-dependent tilt and translation with conjugate eye movements

157

Future work

Head orientation during adaptation during first test during second test rotation axis g vector

components stimulated



Pitch upright VT & NO Roll onside VT (&IA) Yaw onside NO (&IA) Roll upright VT & IA Pitch supine VT (&NO) Yaw supine IA (&NO) Yaw supine NO & IA Pitch upright NO (&VT) Roll upright IA (&VT)

158

Acknowledgements

Advisor: Dr. Mark Shelhamer Committee members: Drs. Paul Fuchs, David Solomon, David

Zee, Kechen Zhang Collaborators: Dr. Stefano Ramat, Dr. Dominik Straumann,

Zurich University Hospital students & staff Technical assistance: Adrian Lasker, Dale Roberts,

NASA staff, Dr. Andy Clarke Scientific advise: Dr. Mark Walker, Dr. Howard Ying, SPH

Biostats Clinic Undergraduate researchers: Ondrej Juhasz, Anton Aboukhalil,

Tiffany Chen, Michelle Zwernemann Moral support: Zee lab, friends & family Funding: NSBRI, NIH, NSERC

160

Control of motion and posture

(AP Photo/Misha Japaridze)



161

The Vestibular System “Sixth sense” that keeps us balanced Stops us from falling when we stumble Helps move eyes (“gaze stabilization”)

Patients with Vestibular Disease or stroke

Astronauts affected by microgravity

VertigoDifficulty walkingEyes move incorrectly

Vertigo and motion sicknessDifficulty walking upon returnEyes move incorrectly

Studying astronauts will help us cure people on the ground

162

Background & Significance Astronauts and others exposed to unusual

acceleration environments get sick Torsional misalignment found in parabolic flight and

may be due to otolith asymmetry Motion sickness correlated with torsional misalignment Motion sickness correlated with an otolith mass

asymmetry in fish In pitch head movements, the otoliths detect a

changing g vector and contribute to eye movements

163


Otoliths measure linear acceleration and gravity (contains utricle and saccule)



Torsion is rotation of eye about the line of sight



Eye misalignment between L and R eye positionVisual disparity between what is seen by L and R eye

164

The otoliths: a mass on a lever

The lever is a hair cell that measures deflection

The mass is called the otoconia and deflects when acted upon by external forces

Gravito-inertial acceleration (g level): the sum of linear acceleration and gravity. 1 g of downward gravity is

indistinguishable from 1 g of upward acceleration of the head

Temporal bone of the

head

165

Otolith Asymmetry Hypothesis

Left Otolith

Right Otolith

Left Eye

Right Eye

g level (GIA)

Central Compensation

166

Implications of otolith asymmetry1. It can predict space sickness: a better understanding

of the mechanisms may help to improve screening and produce simpler screening tests

2. It may reduce performance by misaligning the eyes3. Understanding how it adapts may be useful in

training including partial adaptation before flight

169

Hypotheses An otolith asymmetry will manifest as ocular

misalignments when the magnitude and orientation of the g vector is unusual.

An otolith asymmetry will imbalance the otolith contribution to the pitch VOR, resulting in ocular misalignment

A reduction is misalignment will occur with experience in an environment, and this adaptation occurs within the central compensation mechanism

172

Experimental Methods: Parabolic Flight

173

Parabolic Flight Trajectory

175

Experimental Methods: Video eye movement recording Binocular (both eyes) 50 Hz Accelerometers and

rate sensors Subjects in darkness Software finds pupil in

image and computes gaze direction

178

Experimental Methods: Summary

0 20 40 60 80 100 120-4

-2

0

2

4

Par

abol

ic fl

ight

0 20 40 60 80 100 120-4

-2

0

2

4

Cen

trifu

gatio

n

0 20 40 60 80 100 120-4

-2

0

2

4

Pitc

h ro

tatio

n

Time (seconds)

Parabolic flightOtolith: vertical; change in magnitude (0-1.8 g) but not directionPitch rotation: 3 º/sRoll rotation: 0 º/sYaw rotation: 0 º/sCentrifugationOtolith: rotates in roll to vertical; change in magnitude 1-2 g and directionPitch rotation: 0 º/sRoll rotation: 1 º/sYaw rotation: 0 º/s to 100 º/s

Pitch rotationOtolith: continuously rotating in pitch; fixed magnitude gPitch rotation: up to 90 º/sRoll rotation: 0 º/sYaw rotation: 0 º/s

180

Perceptual Observation Operator: “Did you notice the light diverging at all?” Subject: “The little light… diverged, and I couldn’t get it to

come back together again when I was looking off to the right. The two divergent red lights were not always in the same relation to each other.”

Operator: “Did they separate completely horizontally, completely vertically or something in between?”

Subject: “…It was mostly vertically, but the bottom one would move across, moving horizontally more than the top one”

Subject: “I tried to focus them back on top of each other. They were always vertically separated but I tried to get them horizontally aligned.”

Subject #2: “For right targets, one moved up and to the left. For left targets one moved down and to the right.”

Subject #2: “At the end I didn’t notice it as much at all.”

181

Vertical Results: Video0 1 2

Gravity

184

Vertical eye position trace

187

Algorithm Automatically select a smaller rectangular landmark Find in each video frame using cross-correlation “Temporal feature-selection”

Goal Develop a method to detect vertical translation of the camera relative to the eye using features in the video image… using an automatically selected landmark

189

Co-correlation Metric to estimate motion of a landmark relative to other landmarks

Vertic

al loc

ation

of la

ndma

rk

Time (seconds)

194

3: How is disconjugacy influenced by target distance and position?

Target Near (12 cm) Far (30 cm) Overall

Right +1.54°±1.13° (n=18) +0.99°±1.04° (n=4) +1.44°±1.16° (n=22)

Left +0.99°±0.63° (n=9) +0.70°±0.65° (n=3) +0.92°±0.62° (n=12)

Center +0.90°±0.67° (n=9) +1.15°±0.00° (n=1) +0.93°±0.63° (n=10)

Up +1.20°±0.91° (n=4) - (n=0) +1.20°±0.91° (n=4)

Overall +1.24°±0.93° (n=40) +0.90°±0.79° (n=8) +1.18°±0.93°(n 48)

The magnitude of g-dependent vertical skew does not depend on the horizontal or vertical displacement of the fixation target (ANOVA; p>0.4)

Vertical skew is significantly smaller for far targets compared to near targets (t-test; p<0.1)

Aim 1: Parabolic flight vertical

196

Adaptation Vs. flight experience

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

Parabolic flight experience in current campaign (flights)

Mag

nitu

de o

f ske

w d

iffer

entia

lbet

wee

n 0

g an

d 1.

8 g(

degr

ees)

Influence of flight experience on magnitude of skew differential

Significant trial (p<0.005)Not significantExponential fit (τ=3.69 flights, p<0.034)

198

Adaptation (Torsional disconjugacy)

201

Summary Aim 1: Determine how binocular alignment is disrupted during linear acceleration

The results provide evidence for vertical skew related to g level, possibly as a consequence of otolith asymmetry

The skew is does not varies with target position (comitant) but reduces with target distance

Vertical skew and torsional disconjugacy is reduced with exposure to parabolic flight

The relationship between vertical and torsional disconjugacies will be studied

203

Methods Rotate full body

in the pitch direction

Slow, medium, fast paradigms

Eye movements recorded with scleral search coils

Aim 2: Pitch Rotations

0 20 40 60 80 100 1200

100

200

300

Slo

w(6

deg

/s)

Position (deg)Velocity (deg/s)

0 20 40 60 80 100 1200

100

200

300M

ediu

m(6

0 de

g/s)

0 20 40 60 80 100 1200

100

200

300

Fast

ste

ps(9

0 de

g/s/

s)

Time (seconds)

204

0 20 40 60 80 100 120-10

0

10

Tors

iona

l eye

posi

tions

(deg

rees

)

0 20 40 60 80 100 120

0

2

4

6

8Vertical skew during slow pitch rotation (6 deg/sec)

Ver

tical

eye

posi

tions

(deg

rees

) Right eyeLeft eye

0 20 40 60 80 100 1200

100

200

300

400

BackwardRotation

ForwardRotation

Cha

ir po

sitio

n (d

egre

es)

Time

Slow pitch rotation

205

Medium pitch rotation

Pitch angle (deg)

Verti

cal s

kew

(deg

)

60 º/sec

Target flashes every 7 seconds resetting skew

Skew changes by 0.94º.

Will attempt to reduce noise by finding and removing fast phases

210

Summary of Aim 2: Determine how binocular alignment is disrupted during dynamic tilt (pitch VOR)

Vertical skew occur in pitching with different motion profiles

Torsional disconjugacy occurs in slow pitching; will look in medium, fast

Will look for correlation between vertical/torsional Will perform regression of all experimental data to

determine dependency on motion variables

215

Can an anatomically-based otolith asymmetry model explain the vertical and torsional disconjugacies?

Does a model suggest that the pathway is a direct otolith-vertical or that it is otolith-torsion-vertical?

Does the model correctly predict changes in disconjugacy with different vestibular and visual inputs?

Can the model predict adaptation? How are adaptation and central compensation related?

Aim 4: Model how otolith asymmetry could contribute to disruption of binocular alignment under these different motion scenarios

220

B: Does a model suggest that the pathway is:direct otolith-vertical or otolith-torsion-vertical?

A direct pathway is suggested by evidence that damage to the otolith pathway will result in vertical skew.

Otolith asymmetry

Torsion

Vertical disconjugate and conjugate offset

?

Torsion about an axis other than an optical axis will result in vertical movement.

It has been shown that torsion results from otolith asymmetry.

The model incorporated the geometry of how torsion about different axes would cause vertical movements. We will use experimental data to determine which axis is consistent

225

D: How to model influence of central compensation on vestibular nuclei


+

Vestibular nuclei

Left otolith

Right otolith

Left eyeRight eye

—


Left otolith

Right otolith

Left eyeRight eye

—

X

X

226

D: How to model influence of central compensation on vestibular nuclei


Left otolith

Right otolith

Left eyeRight eye

—

f

f

Context-specific adaptation

229

Acknowledgements

Advisor: Dr. Mark Shelhamer Parabolic flight studies: Dr. Stefano Ramat, Ondrej

Juhasz, Michelle Zwernemann, Anton Aboukhalil, Tiffany Chen, NASA staff

Pitch rotation: Dr. Dominik Straumann, Dr. Stefano Ramat, Zurich University Hospital staff

Scientific insights / moral support: Dr. David Zee & the lab

230

Recruitment The lab is looking for new students – rotations,

Masters, Ph.D.s Mark Shelhamer

[email protected]

232

Otolith asymmetry: Implications for disruption of human binocular alignment in

unnatural gravito-inertial environments

Faisal Karmali

October 11, 2006

An Agenda for Sensorimotor Research in Sub-Orbital Flight · Experienced Day 1 Day 2 Day 3 0 0.5 1...

Documents

Transcript of An Agenda for Sensorimotor Research in Sub-Orbital Flight · Experienced Day 1 Day 2 Day 3 0 0.5 1...