Sensorimotor Adaptation –

Vestibular and Oculomotor


with applications to space flight

and an excursion into fractals


Research in my group is in two main areas:

  • sensorimotor adaptation to space flight
  • fractal statistics in physiology


All of these projects have as their general aim the understanding of sensory processing for motor control, with an emphasis on adaptive capabilities and mathematical modeling.


The  laboratory (part of Dr. David Zee's laboratory in the Department of Neurology) is also associated with the vestibular clinic of the Johns Hopkins Hospital. 

Research is supported by grants from NIH, NSF, and NASA, and previously (and hopefully in the future) by the National Space Biomedical Research Institute (NSBRI).

Sensorimotor Adaptation to Spaceflight


Sensorimotor assessment procedures evaluated in parabolic flight – vestibulo-ocular test (above), posture test (below).


Long-duration spaceflight (~6 months on ISS) leads to a number of physiological problems in astronauts: bone and muscle loss, cardiovascular deconditioning, psycho-social effects, radiation exposure. Many of these areas are being explored by colleagues at JHMI. As part of our general interest in adaptation to spaceflight, we are developing equipment and procedures to perform a rapid sensorimotor evaluation on an unassisted astronaut who has landed in a remote area (asteroid, moon, Mars, Russian steppes). This will allow us to obtain information on the effects of adaptation to spaceflight very soon after landing. This information can then be used to assist in the design of countermeasures, by helping to determine which sensorimotor alterations are the most serious and what the timeline of recovery is. Another application is for an astronaut, perhaps having just landed on Mars, to obtain an assessment that indicates whether he or she is ready to put on a spacesuit and leave the spacecraft, or if there should be a waiting period to allow adaptation to the gravity environment. Sensorimotor tests included in this assessment are vestibulo-ocular function, spatial perception, posture, and locomotion. Given NASA’s love of acronyms, we call this SARA: Sensorimotor Assessment and Rehabilitation Apparatus. (NASA project, with graduate student Kara Beaton)

Context-Specific Adaptation


MATLAB Handle Graphics

Example of improved contextual adaptation when short rest intervals are interspersed between adaptation exposures, for saccade adaptation (above). Example of contextual adaptation of the VOR in different rooms, with VOR gain associated with the room in which it is imposed (below). Effect is stronger when the rooms have less in common with each other.


One form of adaptation to space flight is context-specific adaptation (CSA): the acquisition of two different adaptive responses at the same time, with switching between them based on a context cue. For spaceflight, such a context cue might be gravity level – the brain must process sensory information differently in different gravity levels, and this typically takes a few hours or days of adaptation; CSA would allow immediate switching between the appropriate adapted states in each g level. In the lab, we study CSA by adapting the amplitude of saccadic eye movements (rapid eye movements made to visual targets) or the gain of the vestibulo-ocular reflex (the compensatory eye movement that maintains steady gaze when the head moves). Some property of the response is adaptively altered in one way (e.g., gain increase) with, for example, the head in one orientation, and altered in another way (e.g., decrease) with the head in a different orientation. After adaptation, the particular response that is invoked becomes dependent on the instantaneous head orientation. We are exploring the rate of acquisition and the limits of this adaptive capability, and have recently shown that the setting in which adaptation takes places (the specific room) can serve as a context, such that different adapted states can be present in different places. (Previously supported by NSBRI)

Commercial Spaceflight


Virgin Galactic spacecraft in aerial flight with mother ship (left). Craft will fly a suborbital trajectory, producing weightlessness (free fall) for 4-5 minutes. Centrifuge facility at NASTAR Center (right), used for training in preparation for suborbital flight. Centrifuge has 10 g/sec capability, fully gimbaled cabin.

It is not only long-duration space flight that leads to sensorimotor problems. Short flights can lead to motion sickness and disorientation. This is not a problem for NASA, with their emphasis on long stays on ISS and future trips to near-Earth objects. However, several private companies are preparing spacecraft for commercial spaceflight operations – initially space tourism, and hopefully space research. An immediate research area is the prediction and prevention of motion sickness on these flights, not only to make the passengers more comfortable on their expensive journey, but to enhance the productivity of researchers on the flights. (For information on commercial spaceflight see the Commercial Spaceflight Federation, and the Next-Generation Suborbital Researchers Conference sponsored by the Suborbital Applications Researchers Group)






Sensorimotor Prediction and Adaptation


Initial finding of relationship between fractal structure in a prediction task (abscissa) and rate of adaptation (ordinate). [Figure reduced in resolution until findings can be conformed.]

We have been studying the relationship between adaptability and variability in human eye movements, with the aim of predicting adaptive capabilities in different sensorimotor tasks. A related aim is to find strategies for enhancing the adaptation process. This work has implications for physical rehabilitation, in which knowledge of the ability of a patient to adapt to a specific problem might be used to tailor rehabilitation exercises to favor those areas that are more susceptible to adaptive alteration. We have recently found a relationship between predictive eye movements and adaptation ability. (NIH project, with graduate student Aaron Wong)

Fractal Properties of

Vestibular Afferents



Initial finding of fractal structure (1/f form of spectrum) in a regular afferent of the vestibular system. Fractal structure ends at ~0.1 Hz, beyond the temporal range of natural head movements.

We recently identified fractal properties in resting rates of vestibular afferents in mice. Fractal structure implies activity correlated across long time scales, leading to long-range dependence in the time domain and power-law decay of the frequency spectrum. In vestibular afferents, fractal structure only exists over time courses greater than those of normal head movements (more than a few seconds), which suggests that fractal structure (temporal correlations) assists behaviors that occur over longer time courses, such as velocity storage and bilateral balance, and compensation for aging and other gradual changes. There are important implications for vestibular prostheses, which might benefit from incorporation of fractal firing patterns in presenting head-movement information to intact afferents. (With David Lasker, Charles Della Santina, and Steve Lowen)

Assessing Fractal Structure

in Physiological Signals


Fractal pattern of state-space trajectories for a reflexive eye-movement record (OKN).

Many computational tools to study fractals have been developed. The large number of tools makes it difficult to make a proper choice when analyzing data: some give erroneous or misleading results when improperly applied or when used on data for which they have not been designed. We are assembling a set of established computational tools for fractal time-series analysis, and characterizing them with application to a variety of data sets with known properties. Our goal is to determine cases in which different tools break down, where they are most effective, how to detect erroneous results, and how to interpret the results. (NSF project, with Steve Lowen of Harvard Medical School) 








Student Research