Thomas J. Riess, D.P.M.

176 Morningside Dr.
San Anselmo, CA 94960
(415) 457-8961

Abstract: A normal model of gait is proposed followed by a model of Parkinsonian gait with the goal of construction of a gait enabling therapeutic device. The fundamental underlying tenet of the model is that vision pathology is responsible for the majority of Parkinsonian gait pathology. Several gait-enabling devices are described, an augmented reality one.

1. Introduction

The original motivation for this investigation was the desire to construct vision-based therapeutic devices designed to overcome gait problems associated with PD. The basis for such an attempt is the well documented phenomenon known as Kinesia Paradoxa, whereby in the presence of certain so-called visual cues a PD subject can be transformed from a totally immobile, helpless victim of this disease into a near normal walking individual. An effective vision-based device has indeed been constructed, although the device needs to be further perfected and refined. In the course of developing such a device it quickly became apparent that there was a need for a model of Parkinsonian gait in order to identify and quantify the parameters involved in gait enablement. This model, described below, undergoes constant revision and although it remains essentially unproven speculation it has proven invaluable as a predictive tool of how Parkinsonians react in different and novel environments. The ultimate test of any such model is its ability to effectively define new and better answers. Indeed, as the model evolved, new and different approaches to gait enablement were predicted and on construction proved effective. Several of these prototypes are described at the end of this paper. It should also be noted that since this model is based on observing the phenomenon rather than neurophysiology, neurologic terms have been redefined with slight yet significant differences.


2. A General Model of Gait

Gait can be thought of as learned motor activity of the lower extremities which when performed results in movement of the subject in the environment. By varying the various parameters of gait, (stride length, velocity, heel and toe strike timing etc.) one can perform an infinite variety of gaits. Like any other learned information (motor or sensory or cognitive) a particular gait can be newly learned or overlearned. A newly learned gait exists in short term memory. Characteristics of a given environment determine the particular characteristics of a newly learned gait. Once the environment changes, the newly learned gait is no longer needed and like short-term memory is quickly forgotten and replaced by a different newly learned gait tailored to new environmental conditions. While the environment does vary, there are a few environmental conditions that recur repeatedly and are present most of the time. This results in repeated learning of a select few newly learned gaits until the learned information is transformed in the brain into what is termed overlearned information. Overlearned information differs from newly learned information in several important ways. The most important distinction is that once initiated, an overlearned motor package can be performed unconsciously or automatically. This allows our conscious attention to be directed to another task and is the basis for efficient simultaneous task performance. I would suggest that this is mediated by the basal ganglia and can be thought of as its evolutionary purpose.

Newly learned gait is defined or cued by the external environment (for example, fording a stream by stepping on randomly protruding stones) and is ephemeral whereas the overlearned motor package is predefined and permanently stored in the brain. An internal cue prompts the performance of the entire overlearned motor program as a unit but I would suggest that the essential parameter, which triggers this internal cue, is what I call minimal threshold stride length. For example, if one physically moves the lower extremity of a totally akinetic PD person beyond the minimal threshold stride length he will be able to complete at least one step. The following concepts of stride length are, I believe, part of normal gait function.

1. Initiating Stride: a consciously taken stride, which always follows a position where the feet are separated by less than a threshold stride length stride. The length of the stride is defined by environmental conditions but the stride itself is under conscious control. It is possible to have a series of initiating steps or it is possible to have a single initiating step followed by unconscious automatic gait. In the absence of pathology the environment determines which option occurs.

2. Threshold Stride Length: the minimal inter-pedal distance required to access automatic gait. When this minimum is reached the stride serves as a template for access and performance of automatic gait, which will occur if the environment is compatible with automatic gait.

3. Sub Threshold Stride Length: a stride where the inter-pedal distance is less than that of a Threshold Stride Length. A sub-threshold stride length can be as little as no stridelength whatsoever (for example, when one is standing still with feet side by side) to just slightly less than the threshold stridelength. In normal people environmental conditions, which result in stride lengths, which fall below this minimal threshold, result in externally cued, consciously controlled gait. A conscious, initiating stride will always follow a sub threshold stride length stride.

To summarize, gait can be performed in two different modes; conscious mode and unconscious (automatic) mode. Conscious, newly learned gait is externally cued and its parameters change constantly to adapt to the changing conditions of the environment. Automatic gait is prerecorded or predefined. An externally cued stride serves as a template and when certain critical parameters match an automatic (overlearned) motor package that package is accessed.

Whenever the environment permits we function in the automatic mode (it can be thought of as the default mode). This allows us to perform more complex tasks by assembling overlearned motor packages. When the environment is incompatible with the automatic mode we function in the conscious mode and then revert to the automatic mode when the environment allows it.

Sensory information is required to evaluate the environment and make the determination that a specific program can be performed under the current environmental conditions. The appropriate mode (overlearned versus externally cued) is selected as a result of a dynamic interaction between the environment and the brain. This interaction is mediated via sensory information, which is perceived primarily (but not exclusively) through the sensory organ known as the eye.

The function of sensory information derived from subject - environment interactions is to acquire information about the external environment (environment based sensory information) and to acquire information about how the environment responds as we interact with it (subject induced sensory information). The stimuli resulting in the former originate in the environment while the latter are dependent on the subject for their existence. In regards to motor function environment based sensory information is required to monitor the environment to assess its compatibility with specific motor activity i.e. it defines the motor activity. Subject induced sensory information directly influences the characteristics of the motor activity being performed (if possible, what you see is what you get). Thus, the processing of visual data can be functionally organized into two distinct categories, which have evolved in response to two types of visually perceived sensory information.

1. conscious vision (seeing): visually sensed data plus conscious processing. The data perceived through conscious vision only becomes meaningful after it has been analyzed on a higher cognitive level - filtered through the eye of experience. For example identifying an object in a room as a chair or recognizing a face.

2. unconscious vision: visually sensed data which does not require conscious interpretation, it is processed without conscious awareness. Unconscious vision takes place in the presence of visual data, which is consistent with the motor activity of which it is a consequence (e.g. the apparent motion of the environment as a subject walks through it). Non-conforming visual data is the prompt, which results in a shift from unconscious to conscious vision and results in the subject's consciously altering his motor function to adapt to a perceived change in the environment.


3. A Proposed Model For PD Gait

In PD I would propose that the underlying cause of the majority of gait problems is impairment in the visual processing of motion data. Specifically, this impairment is an inability to properly process motion, which falls below a certain velocity, and small changes in the velocity of motion i.e. acceleration and deceleration. This may in turn manifest itself as (or be a consequence of) an impairment in depth perception (data which originates at greater distances from the observer will appear to exhibit lower velocity optical flow and lower rate of expansion). This impairment is the underlying pathology of all gait problems associated with PD although the presence of what I call the T-factor can make some of these gait problems even more malignant and will be discussed shortly. This impairment spans a range of gait problems ranging from the dopamine deficient state (e.g. akinesia) to those of the fully medicated state (e.g. dyskinesia). Indeed, I would suggest that akinesia and dyskinesia are manifestations of the same underlying visual pathology. The difference is the former is being expressed in the absence of dopamine and the latter in the presence of dopamine. This theory is based on the discovery that visually accelerating optical flow above a certain minimal threshold can overcome akinesia, festination, freezing and dyskinetic gait. There are numerous ways to create environments where optical flow is accelerated. For example walking on a moving sidewalk, roller-skating, bicycling or any of a number of total immersion virtual reality scenarios. PD subjects who exhibit ambulatory dyskinesia will frequently speed up their gait (i.e. accelerate their optical flow) to facilitate the suppression of dyskinetic gait. Also, note that running overcomes dyskinesia for the same reason. And finally walking over regularly spaced objects, the classic scenario described in kinesia paradoxa, also serves to accelerate optical flow. In fact it is not the "objectness" of walking over an array of objects which enables gait but rather their role as reference markers on the real world whose apparent motion serves to augment the visibility of optical flow as we walk over them. This is analogous to watching the dashed lines on a highway "move towards us" as we drive over them thereby enhancing the visibility of optical flow.

When undermedicated this visual pathology is perceived as an inability to generate the required weight-bearing stride length for a particular gait the environment requires at any given time. When undermedicated the only available externally cued stride length (if any) is the very small 2-3" stride length associated with festinating gait. In order to enable gait which accelerates from a standstill to cruising speed one must be able to process optical expansion (the perception that objects appear to enlarge as we approach them) at low velocities and through small changes in velocity. Since this ability is compromised in PD, gait initiation problems occur because stride lengths fall short of the demands of the environment. For the same reason the self-cueing automatic gait program, (though undamaged) is inaccessible, in spite of the fact that the environment is potentially compatible with its performance. This is because conscious stride lengths (if any) do not approach the stride length parameters necessary to create the template to switch it on. Even if the automatic gait package were accessed, it would break down because the same pathology exists in the peripheral processing of data which is essential for sustaining the automatic gait package. In the absence of normal peripheral processing peripherally perceived data will seem inconsistent with what the brain would expect relative to the motor activity it is performing. Thus, the automatic motor program cannot be sustained, and the result is the freezing associated with a state of undermedication.

Consistent with this hypothesis is the observation that activities, which result in accelerated or augmented optical flow immediately, suppress dyskinesia. Examples include bicycling, skating, skateboarding, skiing (including riding on a chair lift), running, walking on a moving sidewalk. In summary, conscious externally cued gait (e.g. when trying to initiate gait) requires the perception of the expansion component of optical flow or augmented centrally perceived optical flow of virtual motion. In its absence the result is akinesia.

The PD subject is often thought of as being more dependent than normals on visual feedback of the self-perception of motion for initiating and sustaining ambulation. I would suggest that this dependency on optical flow is the same in normal and PD subjects. What differs is the ability to perceive it. Normally, when gait is internally cued we monitor the optical flow of the environment (virtual motion) visually but below the level of conscious awareness, in a manner similar to the way we can monitor the subtle motion of being in a boat on the ocean i.e. below the level of conscious awareness. (Nevertheless, such sensory input can impact on our motor skills - e.g. motion sickness). Freezing is the pathological inability to sustain unconscious, automatic gait in spite of the environment being compatible with automatic gait and the subject’s intention to achieve it. Freezing occurs when there is insufficient, erroneous, or misinterpreted visual perception of the optical flow generated by ambulation. Once frozen the subject is forced to change to an externally cued gait. But very often this same pathology manifests itself as an inability to produce an initiating stride of adequate length to conform to the demands of the environment. This is what I would call gait initiation problems which can be defined as the inability to produce an initiating stride of adequate length to conform to the demands of the environment – a failure to meet one’s conscious expectations of stride length. Consider the following analogy. You are stopped at a red light at an intersection on a steep incline. The parked car to your right begins to back up. You "see" this peripherally but initially your brain perceives this as apparent (not real) motion and concludes that you are moving forward. The resulting conflict between what the brain is predicting should happen based on the motor program it is running (foot on the brake) and the accompanying visual feedback results in a kind of reflexive slamming on the brakes. The brain cannot produce motor activity, which is inconsistent with predicted visual feedback. Sustaining unconscious, automatic gait requires the processing of virtual motion (optical flow) which is generally perceived via peripheral vision. Whenever peripheral optical flow disappears or appears to disappear freezing results. For example, doorways frequently induce freezing in PD subjects, but I would argue it is not the doorway but rather the opaque walls around the doorway which precipitate freezing by obstructing one's ability to perceive optical flow.

If one is already in motion and the environment requires a change in velocity the result would be festination in the presence of a need to decelerate and freezing if acceleration is required. Festination occurs whenever we resist an accelerating force imposed on us by the environment (for example, going downhill or maneuvering around an obstacle in our path) and our maximum externally cued stride length is on the order of 2 to 3 inches. To stop (i.e. neutralize the force of acceleration, and a center of gravity displaced forward beyond the fall point) requires that one take many of these small steps in a very short period of time to avoid a fall. If the environment compels one to accelerate, one must elongate one’s stride within the same time interval as previous strides. With only a 2 or 3 inch externally cued stride available this becomes an impossible maneuver and results in freezing. Taking a step requires a deliberate forward displacement of one's center of gravity, which puts one in peril of falling. The perceived availability of a subsequent stride of adequate length to catch up to this displaced center of gravity and save us from falling is what gives us the confidence to place ourselves in this unbalanced position. Conversely if one perceives an inability to extend the leg (either far enough or quickly enough) the fear of falling inhibits the attempt at ambulation. Note how getting up from a chair requires a similar displacement of one's center of gravity. The perceived inability to adequately extend the leg to catch oneself results in consistent undershooting of the forward propulsion needed to get up from the chair (except in the presence of the weightbearing pathologic tone – see "T-factor" discussion to follow). Akinetic subjects have little difficulty negotiating stairs. Again, this is because walking up/down stairs is really walking with greatly reduced forward displacement of one's center of gravity and as a result it is ambulating with minimal stride length (assuming stride length is defined as the length of a stride in the horizontal axis). Also, walking backwards is often possible when akinetic. This is a consequence of human anatomy. Anatomically, walking backwards can take place without posterior displacement of one' s center of gravity.

Motion, both real and virtual, can be perceived centrally or peripherally. Virtual motion can be perceived without conscious attention and this is the basis by which unattended, automatic gait is sustained. When motion is perceived the brain makes a best guess determination based on experience (i.e. sensory conditioning) whether that motion is real or virtual. The ability to distinguish these two categories of motion is essential to the normal performance of gait. Under certain circumstances real motion can be perceived as virtual motion or vice versa. Indeed, such perceptive "mistakes" can be the basis of both gait pathology and therapeutic devices. Sensory conditioning has taught us which characteristics of perceived motion are most reliable indicators of whether that perceived motion should be functionally interpreted as real or virtual motion. Motion perceived as real can define or select a gait motor package. Motion perceived as virtual sustains the performance and influences the quality of performance of that motor program. Dyskinetic gait is an example of what happens when the brain tries to perform an automatic gait package in the presence of impaired perception or processing of virtual motion.

The visual perception of self-motion requires the perception of virtual motion to be perceived. The characteristics of the perception of self-motion (e.g. is the motion fluid or disjointed) are a visual reflection of the characteristics of the motor program being performed. As long as this reflection is consistent with what experience has conditioned us to expect then performance of the motor package is normal. If this visual reflection is at odds with what experience has taught us is the appropriate feedback then this conflict must be resolved either by adjusting the performance of the motor program to conform to the feedback or if this is physically impossible the program stops. This is why monitoring the environment through the viewfinder of a video camera with image stabilization technology suppresses dyskinetic gait.

PD subjects with dyskinetic gait often have a diminished awareness of the degree of their dyskinesia. I would suggest that this is because their motor behavior is in compliance with their aberrant perceptions. It is only when their motor behavior is more grossly abnormal that they become more fully aware of the degree of variance from the norm (perhaps via other modes of sensory feedback such as proprioception or plantar tactile feedback). An analogous situation might be the person who is moderately intoxicated and insists that his behavior has been unaffected.

In spite of normal motor function the manner in which this perceived motion behaves is also a function of how it is referenced to the subject. In the absence of pathology apparent motion of a subject walking at constant speed will appear to be of constant velocity if it is monitored peripherally. In the presence of reference markers on the floor, apparent motion can be tracked in the central field of view. Looking at an outstretched hand while walking (head referencing) results in apparent motion which exhibits constant velocity (the head moves with constant velocity in gait) whereas if one looks at one's feet apparent motion starts, accelerates, decelerates and stops as each foot goes through the swing and stance phase of gait. Both behavior patterns, though different, are normal and expected with normal gait. Deviation from expected behavior of this apparent motion can be a consequence of many and varied causes (for example, orthopedic problems, alcohol intoxication, environmental forces). This model proposes that in PD there is an impairment in seeing or processing low velocity motion or small changes in velocity of motion. If this is the case then one would predict different gait behavior when apparent motion is head referenced (for example walking while looking at outstretched hands (where apparent motion velocity is constant) versus walking while looking at ones feet (where apparent motion varies and motion would be perceived abnormally). Indeed, masking of feet and head referencing apparent motion ameliorates dyskinetic gait.

To this point this discussion has been limited to pathology which is a consequence of a vision perception/processing impairment. However, as PD progresses disturbances of the autonomic nervous system result in increased muscle tone. This increased tone results in an increase in contraction in the major flexion muscle groups and is seen as the stooped posture common to so many PD people (simian posture). I call this the T-factor ("T" for increased tone and tension) whose presence comes and goes as a consequence of the combination of environment-induced anxiety and disturbances in the dynamic equilibrium of various neurotransmitters (on/off cycling).

The presence of T-factor seems to be an augmentation of the fight or flight characteristics of the autonomic nervous system and results in a much more malignant form of akinesia and dyskinesia. When muscles undergo isometric contraction (for example as occurs in the lower extremities while maintaining a standing posture) the additive result of T factor tone and isometric muscle tone results in rigidity - which together with vision pathology comprise the principle components of this more malignant form of akinesia and freezing. If such a subject assumes a crawling position then his arms will have posture maintaining isometric tone and the result is rigidity in the upper extremity and an inability to take a "step" with one's arm. The cogwheel rigidity test is a manifestation of the effects of T-factor when additional tone is induced in a non-weightbearing extremity.

In the absence of T-factor, akinesia and related gait pathology are induced by aberrant perceptions of the environment. Appropriate visual manipulations of environmentally induced perceptions can overcome this gait pathology - the PD subject is physically capable of ambulation. It is the perception of the inability to take a step, which effectively inhibits ambulation. However, in the presence of T-factor the subject becomes physically incapable of advancing the lower extremity and manipulating vision pathology will be less effective. If however, one can initiate ambulation it very often can be sustained because the posture maintaining tone of lower extremity muscles is reduced as the extremities cycle through the various phases of gait thereby reducing one of the two components of tone which contribute to rigidity.

One can frequently observe PD subjects overcoming a freeze by momentarily leaning against a wall. This results in a brief period of reduced weightbearing - a transient reduction in tone which facilitates ambulation. At subnormal levels of dopamine T- factor results in severe akinesia. While medicated the result is freezing which occurs in performance anxiety situations (trying to get to a ringing phone or crossing a street with approaching traffic). Note that T-factor pathology is independent of vision based pathology. It is possible to be akinetic and freeze in the absence of T-factor i.e. vision pathology alone for example freezing in a doorway. It is also possible to freeze in the absence of vision pathology, i.e. T-factor alone which occurs while "on" under performance anxiety scenarios.

Vision only

T-factor only

Doorway freeze

Gait dyskinesia

Performance anxiety: crossing a street

Non-weightbearing dyskinesia


4. Definition of Key Terms

externally cued gait (conscious walking): gait where the gait parameters are defined by the environment

internally cued gait (automatic, unconscious walking): gait which is pre-programmed or overlearned

freezing: vision based pathology resulting in the inability to sustain unconscious, automatic gait in spite of the environment being compatible with automatic gait and the subject’s intention to achieve it.

gait initiation failure: the inability to produce an initiating stride of adequate length to conform to the demands of the environment – a failure to meet one’s conscious expectations of stride length.

festination: a pathologic component of conscious gait which results in a stride which is too small to result in smooth, controlled deceleration in the presence of environmental conditions that require a need to decelerate.

conscious cessation of gait: a conscious successful neutralization of a displaced center of gravity can occur during normal gait or festination.

falling: festination, which does not catch up to a displaced center of gravity.

conscious vision: central field vision (seeing)

unconsious vision: peripherally processed motion, perceived without conscious attention

initiating stride: a consciously taken stride, which always follows a position where the feet are separated by less than a threshold stride length stride.

threshold stride length: the minimal inter-pedal distance required to access automatic gait.

sub-threshold stride length: a stride where the inter-pedal distance is less than that of a threshold stride length.

T-factor: pathologically increased muscle tone, an exaggerated response to normal tone inducing stimuli


5. A Brief Discussion of Gait Enabling Devices

The fundamental underlying principle of all gait-enabling devices is their ability to augment the "perceivability" or velocity of low speed optical flow.

Figure 1: Central Field Cueing Device

Figure 2: Optical flow per stride is that segment of the floor, which appears to pass under foot as we advance the foot in gait. Due to vision geometry the optical flow per stride increases as the subject’s eyes elevate relative to the floor. The velocity of the optical flow increases because the increment of time required to take the stride is unchanged.


All of the above have been clinically tested in unblinded, unscientific, informal trials with varying degrees of success. There are in fact numerous other approaches to augmenting the perceived or actual velocity of optical flow. Many of these have been built and successfully tested but were abandoned due to the constraints of social acceptability or cost of development. All of the devices are in prototype form only and are not currently available to the general public. However, a modest amount of financial and engineering resources would yield great dividends in terms of a useful therapeutic gait-enabling device.


6. Conclusions

In conclusion it is the underlying tenet of this paper that most of the gait pathology associated with PD is vision based. Future research should answer such questions as which parameters of the visual process exhibit this pathology. Is it a perceptive problem of the retina and/or a vision-processing problem in the brain? Is there a problem in "seeing" motion, changes in motion, or depth perception?

The entire phenomenon of Kinesia Paradoxa has been regarded as a curious anomaly of PD. It should be thoroughly studied because it is an important clue in understanding this phenomenon.

Fundamental research in neurology cannot be overestimated in its importance for finding solutions to problems of PD and ultimately the cure. However, much can be learned by a phenomenological based investigation, which can potentially provide interim solutions much more rapidly. The ultimate test of any such model is its ability to effectively define new and better answers to the problems of PD. The concepts expressed in the conceptual model described above are only now being subjected to scientific scrutiny and it is hoped that such investigation will result in further revision and refinement. These solutions result in improved quality of life of those currently afflicted with PD and this has been and remains the primary objective of these efforts.



1. Barbato L., Rinalduzzi S., Laurenti M., Ruggieri S., Accornero N., Color VEPs in Parkinson's disease

2. Brotchie, Peter, Robert Iansek, and Malcolm K. Horne. Motor function of the Monkey globus pallidus. Brain,114:1685-1702, 1991.

3. Buttner T., Kuhn W., Klotz P., Steinberg R.,Voss L., Bulgaru D., Przuntek H.. Disturbance of colour perception in Parkinson's disease. J of Neural Transmission - Parkinson's Disease and Dementia section.6(1):11-5, 1993

4. Campbell RJ., Steele JC., Cox TA., Loerzel AJ., Belli M., Belli DD., Kurland LT., Pathologic findings in the retinal pigment epitheliopathy associated with amyotrophic lateral sclerosis parkinsonism/'dementia complex of Guam. Opthalmology 100(1):37-42. 1993 Jan.

5. Gerigiou, N., R. Iansekl, J. L.Bradshaw, J. Phillips, L. Mattingley, J. A. Bradshaw. An evaluation of the role of internal cues in the pathogenesis of parkinsonian hypokinesia. Brain 116, 1575-1587, 1993.

6. Glickstein, M J. Stein. Paradoxical movement in Parkinson's disease. TINS,14,480-482,1991.

7. J.P.Harris, J.E. Calvert, O.T. Phillipson. Processing of spatial contrast in peripheral vision in Parkinson's disease. Brain.115(pt5):1447-57, 1992 Oct

8. Haug, B., Trenkwalder,C., Arden, G., Oertel, Paulus. Visual Thresholds to Low-Contrast, Luminance Contrast Stimuli in Parkinson's Disease. Movement Disorders vol 9, No.5, 1994 pp 563-570

9. Hoover, J., P. Strick, Multiple output channels in the Basal Ganglia. Science 259,1993

10. Institution: UA CNRS 1166 Cognition and Movement, Universite Aix-Marseille II, France. Review article. Funddam Clin Pharmacol 1993;7(8):449-63.

11. Krasilovsky, G. J. Gianutsos. Effect of video feedback on the performance of a weight shifting task in subjects with Parkinsonism and Neurologically intact individuals. Experimental Neurology 113,192-201,1991

12. C.J. Lueck, S. Tanyeri, T.J. Crawford, L. Henderson. Saccadic eye movements in Parkinson's disease: I. Delayed saccades. Quarterly Journal of Experimental Psychology: Human Experimental Psychology. 45(2):193-210, 1992 Aug.

13. Luquin, M.R., O. Scipioni, J.Vaamonde, O. Gershanik, and J.A. Obeso. Levodopa-induced dyskinesias in Parkinson's disease: Clinical and pharmacological classification. Mov Disord, 7:117-124, 1992

14. Malapani, C., B. Pillon, B. Dubois, Y. Agid. Impaired simultaneous cognitive task performance in Parkinson' disease: A dopamine-related dysfunction. Neurology 44:319-326, 1994

15. Marsden, David C., Parkinson's Disease, Chapter 2. pages 57-98. Chapman and Hall Medical, 1990

16. Marsden, David C., and J.D. Parkes, Success and problems in long-term levodopa therapy in Parkinson's Disease. Lancet, 12:345-349, 2977

17. Price MJ Feldman RG Adelberg D. Kayne H. Abnormalities in color vision and contrast sensitivity in Parkinsons' disease. Neurology42(4):887-90, 1992

18. Riess, T., Weghorst, S., Augmented Reality in the Treatment of Parkinson's Disease, Interactive Technology and the New Paradigm for Healthcare IOS Press 1995, 298-302

19 .Spinney, L. Seeing is Unfreezing. New Scientist 2/15/97 pp38-40.

20. Takeda M., Tachibana H., Okuda B., Kawabata K., Sugita M., Event related potential and visual evoked potential in patients with Parkinson's Disease.

For any questions or requests, please contact