Virtual Reality Therapy
VR applications have emerged in the areas of entertainment, education, military
training, physical rehabilitation, and medicine. One area that is yet to be
exploited, however, is in the application of virtual reality technology for
the rehabilitation of persons with brain injury and neurological disorders.
In this regard, VR could serve to revolutionize the study of brain/behavior
relationships as well as produce treatment options unavailable with traditional
methods. Much of the testing focuses on those individuals having experienced
some kind of brain trauma and since the peak age of incidence is in the 15-24
year range (closely followed by the birth to five year old group), many physically
fit young survivors with damaged brains face a life long struggle with a tragic
set of cognitive and functional impairments. In addition to the cost of human
suffering, one estimate places the economic costs in terms of medical care,
rehabilitation, lost work potential, etc. at $48 billion annually. Similar impairments
occur as a result of a wide variety of neurological disorders including Alzheimer's
disease, vascular dementia, cerebral palsy, epilepsy, Parkinson's disease, and
multiple sclerosis. Alzheimer's disease alone has been estimated to afflict
nearly four million Americans. Many others, particularly the young, experience
cognitive and functional impairments due to various developmental disabilities--attention
deficit or hyperactive disorders and learning disabilities. The potential of
VR lies in its ability to produce tests and training environments that can provide
accurate records of targeted responses.
When discussing possible VR applications for Cognitive Rehabilitation (CR), it is helpful to consider an important finding pertaining to preserved memory abilities following brain trauma. A number of studies have shown that in persons with neurological-based memory impairment, certain memory sub-processes often remain relatively intact. Procedural or skill memory is one such cognitive operation. This type of memory ability concerns the capacity to learn rule-based or automatic procedures including motor skills, certain kinds of rule based puzzles, and sequences for running or operating things. Procedural memory can be viewed in contrast to declarative or fact-based memory which is usually more impaired and less amenable to rehabilitative improvement. Additionally, these patients often demonstrate an ability to perform procedural tasks without any recollection of the actual training sessions. This is commonly referred to as implicit memory and its presence is indicative of a preserved ability to process and retain new material without the personas conscious awareness of when or where the learning occurred. These observations provide encouragement for the idea that VR, by way of its interactive and immersive features, could provide training environments which foster cognitive improvement by exploiting the personas preserved procedural abilities. Hence, cognitive processes could be restored via procedures practiced repetitively within an environment which contains functional real-world demands. Whether the person could recall the actual training episodes is irrelevant as long as the learned process or skill is shown to generalize to functional situations. The real challenge would then be to somehow translate difficult declarative tasks into procedural learning activities with the goal being the restoration of the more complex higher reasoning abilities. However, in order to develop these applications, a wide range of more basic issues need to be addressed.
The current status of this field, while provocative, is limited by the small number of studies that have been reported applying this technology to clinical populations. This is to be expected considering the expense of the technology and its relatively recent development. Although much of the work does not involve the use of fully immersive headmounted displays (HMD's), studies reporting PC-based flatscreen approaches are providing valuable information on issues necessary for the reasonable and measured development of this field. Also it may be found that for some applications and populations fully immersive systems may not be feasible or necessary. Initially, VR has been shown to promote learning in unimpaired populations. These reports indicate VR training success in the areas of console operation, building navigation, operation of manufacturing equipment, training support personnel for the Hubble telescope repair, military terrain knowledge, surgical procedures, and complex naval conning maneuvers. Positive initial results have also been reported for the Train to Travel program which is a VR system primarily designed to teach persons with developmental disabilities how to use key routes on the Miami Valley Regional Transit System. The success of this program would be useful as a demonstration that functional abilities could be efficiently and economically trained in a patient population that is often underserved by new technology. Another research group reported that they had commenced the training of basic environmental exploratory skills with a 36 year-old male with frontal lobe dysfunction and encouraging results using non-HMD VR have been reported in students with cognitive impairments due to developmental disabilities. Their findings suggest the advantages of low cost PC-based virtual environments for training functional activities such as supermarket shopping in persons with severe cognitive impairments. The question of whether persons with cognitive impairments can learn in VR settings, however, is still waiting to be fully answered. Certainly a long and productive academic career awaits someone willing to empirically test each of the above proposed VR advantages on different cognitively impaired populations!
Our MR assessment and training system is designed to present, within a VR environment, a target stimulus that consists of a specific configuration of 3-D blocks. After presentation of the target stimuli, the participant is presented with the same set of blocks that need to be rotated to the orientation of the target and then superimposed within it. The participant can manipulate the blocks in virtual space by grasping and moving a block-like cyberprop object that contains a tracking device and provides tactile feedback. The stimulus complexity of the target can be adjustable via changes in number of blocks and configuration, visual field of presentation, and degree and type of rotation required. Response characteristics, such as the time needed to appropriately rotate the blocks and all movement sequences, can be recorded. The system will allow for the hierarchical control of stimulus complexity and the capacity to record and measure all behavior in the assessment settings, with an aim towards restorative cognitive rehabilitation. Immediate performance feedback (variable pitched sound or color changes) as to the effectiveness of the response approach can be programmed into the system in order to help guide the participant in an errorless learning fashion. To address functional assessment and rehabilitation goals in an ecologically valid manner, the block stimuli would then be replaced with images of real life objects. This format would test and train the planned movements needed to meet the requirements in the VR setting, that are typically found in everyday visuospatial problem-solving situations.
At this early stage of VR development, the primary emphasis has been on transferring information learned from a VR environment to the actual real world environment. Support for this claim can first be seen in the related and prior field of aviation simulator research. In a recent report (1995) on theoretical issues concerning transfer of learning from aircraft simulators, it cites a Transfer Effectiveness Ratio in the aviation simulation research of .48. This ratio indicates that for every hour spent in aviation simulator training, one-half hour is saved in actual aircraft training. VR was also shown to foster learning of console operations and this learning was shown to transfer to the actual console. Self-guided VR training for machine operation has also been shown to promote good transfer. In a project conducted at Motorola University which is of particular practical interest, the results indicated positive transfer from a VR factory training program to the actual factory line. The implications of this work for CR and vocational rehabilitation interests are obvious. Finally, in the only study to examine generalization with clinical groups, evidence of positive learning transfer from a non-HMD virtual training setting was found for a group of developmentally disabled students. Students with significant cognitive impairments were trained on a PC-based, non-HMD, virtual system to navigate through and select specific items in a virtual supermarket. In addition to demonstrating good transfer of learning, this study is noteworthy in that it suggests an efficacious approach to increasing the independence of persons with cognitive difficulties for whom a fully immersive HMD strategy may not be practical.
The above cited investigations represent essential first-steps in determining whether VR training can foster transfer of learning to activities of daily living. For persons whose learning abilities may already be challenged due to neurological trauma, this line of research is especially important. The generalization problem has plagued the overall field of cognitive rehabilitation since its inception. This makes it essential that intuitive expectations of positive VR learning transfer be, in fact, supported with quality research. This will be vital in order for the VR approach to be taken seriously in this field.
The goal of virtual reality is to build virtual environments that minimize the learning required to operate within them, but maximize the information yield. In order for persons with cognitive impairments to be in a position to benefit from VR/CR, they must be capable of learning how to navigate within the environment. Many modes of VR navigation (data-gloves, joy sticks, space balls, etc.), while easily mastered by nonimpaired participants, could present problems for those with cognitive difficulties. Even if patients are capable of interacting in a VR system at a basic level, the extra non-automatic cognitive effort required to navigate may serve as a distraction and limit or slow the rehabilitation process. The use of voice recognition technology may serve as a useful alternative to providing for a more naturalistic interface on some training tasks and improving VR access for persons with motor impairments.
The useful application of VR in the areas of assessment and rehabilitation of cognitive functioning, while intuitively appealing, cannot progress until certain basic questions are addressed. Initial pragmatic concerns in determining the feasibility of using VR for these purposes include: differences in patient populations (young head injury patients vs. senior citizens with early indicators for alzheimer's), the potential for VR induced cybersickness in persons with various neurological impairments, and whether persons with impaired cognition can initially learn how to navigate and interact within a VR environment. After these feasibility concerns are addressed, we will be better prepared to ask such questions as, to what degree can these patient groups learn in VR, and does this VR-based learning generalize or transfer to real-world situations. Our research program has been designed so that these issues could be economically addressed, while at the same time, data can be collected regarding our cognitive variable of interest in visuospatial mental rotation.
The building of a foundation for the application of VR techniques in the area of cognitive rehabilitation will require the investigation of a wide range of issues. At this early phase of VR technology, a number of obstacles exist which have impeded the development of active research specifically testing persons with cognitive impairments. These obstacles include the relative lack of familiarity with the technology and problems with funding acquisition for an untested and fairly expensive new treatment modality. On the other hand, there is certainly no shortage of theoretical discussion on the potential value of VR in this area. While the obstacles can be seen as short-term difficulties (VR awareness is increasing while system costs are decreasing), creative research approaches are needed to address initial feasibility questions and manifest VR's potential. Researchers hold strong beliefs that VR offers a unique set of advantages for the delivery of rehabilitative strategies to persons with cognitive impairments. However, the what if questions in our theoretical musings will now need to be replaced with what is answers based on objective studies of the issues.