1998 Conference Proceedings

Go to previous article. 
Go to next article. 
Return to 1998 Conference Table of Contents 

Use Of Virtual Reality Techniques To Train Elderly People To Step Over Obstacles

David L. Jaffe, MS
Research Biomedical Engineer
Department of Veterans Affairs
Palo Alto Health Care System
Rehabilitation Research and Development Center
3801 Miranda Ave., Mail Stop 153
Palo Alto, CA 94304
650/493-5000 ext 6-4480
650/493-4919 fax


This project employs virtual reality techniques to improve the ability of elderly individuals at high risk for falling to step over obstacles.


The long term goal of this work is to construct and test a system for elderly individuals at high risk for falling that 1) employs techniques to monitor and improve their stepping-over response time and 2) trains them in more effective movement strategies. This method is expected to be safer than conventional training and more rapid and precise in the feedback it provides to the patient.

Currently, the only method of training stepping-over responses involves exposing the subjects to the actual hazards, such as practicing stepping over different sized objects. Many activities can not be practiced because they 1) involve too much of a risk, such as stepping-over a moving object, 2) may lead to an actual fall, such as stepping over barriers, or 3) are not possible to simulate, such as having a dog suddenly run in the gait path.

Simulated obstacles of various sizes, shapes, colors, and location in the user's path will be presented via a head mounted display as they walk on a treadmill. Foot height will be monitored to analyze users' movements and detect collisions between the virtual object and the users' feet. This experience should promote effective stepping-over movement strategies.


"Falls and instability are among the most serious problems facing the elderly population. They are major causes of morbidity, mortality, immobility, and premature nursing home placement. A fall also constitutes a marker condition of an underlying reversible medical or psychosocial problem that can be corrected."

Statistics on falls in the US document the serious consequences of the problem. Accidents (a majority of them are falls) are the fifth leading cause of death in those over 65. Death results from 75% of falls in the 12% of the population that are over 65. One percent of those that fall are admitted into a hospital with only half of these surviving after the incident. Approximately 90% of hip fractures in elderly people are caused by falls. Of those elderly people who were functionally independent before a hip fracture, 25% remain in long term care for more than a year afterwards, and another 35% must depend on mechanical aids or people for mobility. On a more positive note, the Public Health Service states that 2/3 of the deaths due to falls are preventable. Therefore, there is a great need to identify the causes of falling in the elderly and develop strategies for preventing falls.

Currently, the preferred training method for reducing fall risk in frail elderly persons involves exposing them to real hazards by practicing stepping up and down curbs or stepping over obstacles placed strategically about the clinic. Many potentially hazardous activities are not practiced because they involve too much of a real risk to the subject, may lead to an actual fall, or are too difficult to simulate in a clinical environment. It is essential to develop training methods that will provide a safe environment, will increase awareness of potentially dangerous conditions, and will help persons to regain confidence after falling.

One potentially safe environment for ambulation training is the treadmill. Overhead harnesses may be employed to further stabilize persons with impaired gait. Also, treadmill training has been shown to have a positive effect on movement capability. Participation in a treadmill training program by patients who had suffered myocardial infarctions increased self-reported activity at home because of reductions in the subject's fear of falling.


Frail elderly persons who demonstrate impaired stepping-over responses can be trained, using computer generated obstacles, to improve their gait speed and to step over objects. This improvement will exceed that typically seen using conventional overground gait-training regimens and with standard (e.g. no computer generated obstacles) treadmill training.


The long term objective of this project is to develop a technique to monitor, train, and improve stepping-over responses (SORs) to typical hazards encountered during walking for elderly persons at risk for falling. To this end, previous work which has demonstrated that computer simulated objects are effective in eliciting appropriate SORs in both young and elderly healthy persons will be built upon. The specific objectives of this three-year project are:

Current methods

There are many problems with the current methods of providing obstacle avoidance training to individuals at high risk for tripping or falling. Practicing these techniques with real obstacles or out-of-doors exposes the patient to the risk of injury due to a fall. Physical therapists report that they limit or don't attempt training to avoid possible injury to the patient.

Simulated environment techniques can address some of the shortcomings of current training methods. With the proposed method, a trainee can make mistakes without fear of injury. This training can occur within a clinical setting. Earlier training may aid in recovery, and improve self-esteem and functionality. For example, treadmill walking has been used as an early intervention in patients with acute CVA. The results of the pilot study showed a 30% increase in walking speed for those in the earlier program as compared to those undergoing later traditional physical therapy. In addition, elderly women who had hip fractures and had participated in a treadmill gait training program demonstrated significantly increased mobility as compared to those who had received traditional gait therapy.

A potential disadvantage of treadmill training is that it can become boring unless suitable visual feedback is incorporated into the simulated environment. This will be addressed by introducing the user to multiple situations that may induce falls and thereby plan strategies for fall avoidance as well as watch for those conditions in the real world. Several studies have identified areas of the home, such as hallways, in which falls often occur and tripping hazards, such as rugs and floor clutter, which can be simulated on the computer. Interesting and informative exercise sessions may include common outdoor hazards such as uneven pavement, curbs, and misplaced gardening equipment. An added benefit of training with interesting visual images is that the number of repetitions of movements increase and motor performance is improved.

Summary of work accomplished

Past work has led to the following conclusions:

Proposed methods

The proposed experiments will demonstrate whether any or all of three training regimens (conventional obstacle training vs. treadmill training with simulated objects vs. treadmill training without simulated objects) are successful in improving gait speed and performance in an obstacle course. The composite image in the head mounted display and walking on the treadmill creates a simulated environment that encourages and improves stepping-over responses that closely match those exhibited during overground walking. To the pilot work real-time foot tracking, "collision" biofeedback stimulators, and a software system that integrates the tracker, video, and stimulators will be added


1. Develop laboratory setup

Images from a video camera and the computer simulation program will be superimposed with a video-mixer and presented to subjects walking on a motorized treadmill wearing a high resolution head mounted display (HMD). Subjects will be wearing an overhead harness to prevent injury in the case of a fall. An obstacle course consisting of 10 objects of various height and length combinations will be fabricated over a 100 foot walkway.

2. Develop virtual display

In this task, the application software will be developed. This will consist of a randomized series of computer generated videos that depict simple rectangular obstacles of varying heights and lengths. Healthy, young and elderly subjects will be used to establish reliability of the system. Subjects will be asked to continuously step over simulated stationary objects displayed at their feet. When the subjects' real foot intersects the computer generated obstacle's virtual position, a "collision" occurs, producing a vibro-tactile stimulus on the appropriate foot. Reliability will be established by videotaping the sessions and visually establishing that the "detected collisions" occur when the foot "touches" the computer-generated object and that collisions are not detected when the foot clears the object.

3. Training sessions

One group of 42 frail elderly persons (65 years of age or older with mild Chronic Obstructive Pulmonary Disease - COPD) will be recruited for this study. Spatio-temporal variables of walking performance will be measured, as will their ability to negotiate a standardized obstacle course with obstacles of various height and length. Those subjects who contact 3 or more obstacles (out of 10) during the obstacle course test will be included. Frail elderly subjects will be randomly distributed to one of three intervention groups - conventional training group vs. simulation-treadmill training group vs. non-simulation treadmill training group. Each group will undergo four weeks of training consisting of three weekly sessions of approximately 60 minutes duration. In the conventional training group, the same trained physical therapist will apply overground gait training techniques to improve a person's capability to step over objects. In the non-simulation treadmill group, a trained physical therapist will train subjects to walk faster, step longer, and step higher while walking on a treadmill. In the simulation-treadmill group, the same trained physical therapist will apply the simulated obstacle training regimen to also improve performance. After 12 training sessions, subjects will again be evaluated using the standardized obstacle course. The outcome measures to be used are: 1) spatio-temporal variables of gait (speed, cadence, and step length) and 2) obstacle contacts during the obstacle course test.

4. Data reduction and analysis

Spatio-temporal variables of walking and SOR performance (i.e., ability to clear obstacles during the obstacle course test) will be compared before and after intervention. Statistical tests will be applied to the data for comparisons of spatio-temporal gait parameters, and obstacle course performance (i.e. speed and number of collisions) in the three intervention groups.

5. Modifications and future work

Videotapes of the training sessions will be reviewed to identify any potential problems with the experimental setup. In addition, feedback from volunteer subjects will be used to assess the realism of the video displays and their responses to the simulated obstacles. During the initial trials, much experience will be gained from working with this laboratory setup, including ideas for improving the experimental procedure. Modifications will be made to improve the quality and realism of the images.


The ability to step over objects is an essential component of ambulation that enables a person to safely function in real world environments. This is especially important for persons with walking deficits secondary to musculoskeletal (amputations, post orthopaedic surgery) and neurological (stroke, head injury, spinal cord injury, multiple sclerosis, peripheral neuropathy, Parkinson's disease) problems. In addition, the frail elderly, a growing Veteran population, would benefit from this research.

The techniques that may develop from this project will encourage repetition and practice in problem areas and the ability to control hazardous situations at a level appropriate to the individual, taking into account his/her progress. The user can get visual feedback of his/her progress. In addition, a quantitative measurement of patient progress will be obtained using a standardized obstacle course.

Current training methods involve practice in walking over objects of different sizes and shapes that are placed on the floor, and is limited by available materials and space constraints. In addition, practice of this task involves a high level of risk to the patient, since failure to clear an obstacle can result in stumbling and falling. This and future proposals can address the need for manipulating important aspects of obstacle avoidance such as obstacle shape, size, fragility, movement, and background environment, while maintaining safety and eliminating material and space constraints. Since this type of training would involve a limitless range of obstacle types, potentially dangerous situations can be assessed safely in a clinical setting. Health costs will be reduced by decreasing the amount of time needed for training for a wide variety of obstacle conditions, and by the potential impact of preventing falls caused by inadequate or inappropriate obstacle avoidance.

Future work

Future work in this project area could include employing this simulation technique with walking aids such as canes and crutches. Other potential areas of research include the study of improvements in fitness and gait through simulation of walking situations for ambulatory nursing home patients and teaching environmental factors and modifications to avoid falls. The system could provide an enjoyable and safe environment for general exercise, a safe setting for "wanderers", or a simulated practice session for wayfinding and familiarization of nursing home patients with their facility.

If this project succeeds in developing obstacle avoidance training methods for elderly people, a partnership with a suitable company to market and provide this system to clinics and medical centers will be pursued.

Go to previous article. 
Go to next article. 
Return to 1998 Conference Table of Contents 
Return to Table of Proceedings 

Reprinted with author(s) permission. Author(s) retain copyright.