Virtual reality training for stroke patients with navigation impairment

Numerous efforts are being made to combine serious gaming and healthcare. Such an approach is also highly relevant to navigation ability. The ability to navigate enables us to find our way around in familiar and unfamiliar places and is considered an important aspect of human cognition. It relies on multiple cognitive processes such as perception, attention, and memory. Such a complex cognitive construct as navigation, based on the involvement of different neural structures and networks, is evidently vulnerable to brain damage.

Former research has established that 29% of stroke patients suffer from impairments in navigation ability. Wayfinding problems are thus common among these patients. This finding is specifically alarming as navigation ability and daily life functioning (i.e. quality of life, mobility, and autonomy) have shown to be positively correlated. We therefore hypothesized that improving patients’ navigational ability might result in an increase of quality of life after mild stroke. There is thus a clear need for interventions (e.g. virtual reality or serious game training) that aim at improving navigation skills in individuals suffering from topographical disorientation.

To investigate whether navigation abilities of brain-damaged patients can be improved, we designed a virtual reality (VR) navigation training program. The purpose of this program was to teach six chronic stroke patients how to apply alternative strategies to compensate for their navigation impairments. Before the start of the training, patients were broadly assessed in their navigation ability to determine the cognitive nature of their navigation impairment. The precise training approach was then specified for each patient separately based on their individual pattern of strengths and weaknesses in navigation ability.

Regarding navigation strategies, a majority of studies point to two fundamentally different strategies based on route knowledge or survey knowledge. Route knowledge refers to information about landmarks, place-direction associations and sequences of landmarks or turns. This type of knowledge is the result of adopting a ground-level perspective. Survey knowledge, in contrast, concerns information of the general layout of the environment and is thought to result in a cognitive map of the area including information about metric distances and angles. This knowledge is typically obtained by extensive exploration of an environment or by map learning.

The training program consisted of three to five one-hour sessions. The content was determined given the individual navigation strengths and weaknesses of the patient. To this end, we used a dynamic virtual rendition of the German city Tübingen which allowed for both free exploration and following routes that were specified beforehand by the trainer. Control of the virtual environment was fully executed by means of a joystick. For example, in case of a patient with impaired route knowledge, the training procedure would be aimed at promoting the patient to use a survey-based strategy. Exercises would, for instance, address the adequate coupling of the ground-perspective of the VR with the map view. For example, patients would be provided with a route specified on a map and were then asked to follow this route through the virtual environment. Feedback was provided by the VR (“turn around”) in case a wrong turn was taken. In another type of exercise, participants had to plan and draw a route on a map and then follow the planned route through the virtual environment. Such an exercise was aimed at promoting the patient to adopt a survey-based strategy. Most of the sessions included the completion of two of such exercises in the virtual environment (using a game-like approach).

Our results offer preliminary evidence that the strategy people use to approach navigational challenges can be influenced or even changed by a relatively simple training procedure. In this study, half of the patients learned to apply a different navigation strategy. Significant improvements in navigation ability on group level were not found, however, in part as a consequence of the small sample size.

In the navigation training, we made extensive use of a dynamic virtual rendition of the German city Tübingen. The feasibility of this virtual environment was found to be high in our training program. All patients were able learn to control it (by making use of a joystick) within a single training session. Virtual Tübingen was considered fairly realistic by the patients as well.

Our findings underline the usefulness of further evaluating the effects of virtual reality navigation training in brain-damaged patients in a more systematic and controlled way. One future possibility would be to develop serious game (navigation strategy training), which first assesses the specific impairments in navigation ability of individuals, then selects an appropriate training program and provides feedback on the performance and progress of the participant. In order to allow for a multitude of different exercises, such a game should include an extensive virtual environment. One major advantage of a navigation strategy training in the form of a serious game would be that patients are able to practice independently at home without the need for the presence of a trainer.