The results from this study indicate that 12 sessions of gaze stabilization exercise using interactive video games over the course of 6 weeks can reduce dizziness, improve balance and increase the walking speed of patients with either BVH or UVH, thereby improving their quality of life. These beneficial effects of training can be maintained for at least 1 month after training completion if complemented with in-home training. Because the interactive rehabilitation sessions are fun and challenging, patients can receive more encouragement and confidence through timely feedback. This goal-orientated, task-specific training method has previously been widely used in patients with chronic spinal cord injury, traumatic brain injury, stroke and cerebral palsy [22, 40, 41].
The major complaints of vestibular patients include imbalance, dizziness, visual confusion or space and motion discomfort (SMD), especially in shopping malls, cars, trains or other places with narrow visual spaces and/or complex and confusing visual stimuli. Virtual reality (VR) technology can provide vestibular dysfunction patients with a controlled environment in which they can gradually acclimate to the scenes that induce their symptoms . Head-mounted displays are easy to carry and affordable, and they were commonly used projection systems for vestibular rehabilitation. However, patients often complain about eye fatigue, blurred vision, headaches, nausea and short-term changes in binocular vision. These adverse effects might stem from changes in the balance between the convergence of the eyeballs [42, 43]. Using a head-unrestrained, wide-field gaze shifts the environment; however, no simulator sickness was reported in healthy subjects . Sparto et al. also used a wide-field visual system to evaluate and train patients with vestibular system dysfunction. It was hypothesized that the peripheral flow is an important sensory recalibration for the patients. It was thus concluded that wide-field VR training may be superior to a head mounted-display for the training of vestibular patients .
Compared with expensive VR technology, the equipment used in our study is relatively simple and lightweight. It does not require a large amount of space to operate and is therefore suitable for patients’ home training.
During everyday life activities, such as running, the speed of head movement can reach 550°/sec, head acceleration can reach up to 6000°/sec2 and the frequency of head rotation can range between 0 and 20 Hz [45, 46]. Given this high speed, acceleration and frequency range, it is difficult for patients with loss of vestibular function to adapt, considering the complexity of adaptation and compensation mechanisms, including sensory/motor substitution and central preprogramming . Smooth pursuit, optokinetic function and cervico-ocular reflexes (CORs) can interact with VOR to reduce retinal slip . Smooth pursuit and optokinetic functions are forms of visual compensation for slow head rotation. Animal experiments have confirmed that the neck and vestibular system can interact to generate visual compensation . However, gains in the COR are very slow and can be difficult to observe in primates . Therefore, the clinical significance of these gains has not yet been confirmed. In our study, in addition to gaze stabilization training, the patients turned their heads frequently during the training exercises, which might have led to improvements in the range of head rotation. Whether this might also minimize dizziness by improving neck proprioception requires further study. With regard to the central preprogramming theory, Herdman  has proposed that visual acuity is higher during active head movements than during passive head movements in patients with unilateral damage. This finding likely results from the fact that when a patient turns his or her head toward the damaged side, the movement is predictable, and the eyeballs can be adjusted through central preprogramming to maintain gaze stability. In this study, the subjects were asked to perform active head movements. Because the head movements of each subject were self-predictable, compensatory eye movements mediated through central preprogramming were able to improve gaze stability and relieve the patient’s symptoms [52–54]. Discussions on compensatory saccade (CS) have recently gained significant attention. This type of saccade can simultaneously occur during both expected and unexpected head movements. It has been suggested that this compensatory mechanism of the non-vestibular system is related to gaze stability improvements [55, 56]. In our study, we did not record the patients’ eye movements; therefore, further investigation is required to determine whether CS has a role in this study.
Our results indicate that the improvement of gaze stability leads to a decreased likelihood of patients experiencing dizziness in response to head movements. The primary functions of the vestibular system are to detect head movement, maintain the stability of images projected on the retinal fovea and maintain postural control during head movements . The vestibule is located in the inner ear, and it functions by detecting the position and movement of the head and by providing the appropriate sensory information to the central nervous system (CNS). Sensory input is primarily sent to the vestibular nuclei for processing and to the cerebellum for the micro-regulation and processing of body balance and coordination. The CNS can then stabilize the head and body via the neural reflexes of the vestibular system. Typically, the CNS maintains the center of gravity of the body between the left and the right foot and takes appropriate actions to maintain body balance. During head movement, the maintenance of body stability relies increasingly on the vestibular system . In our study, the patients were better able to maintain body balance after training. It could be speculated that the vestibulo-spinal reflex was exercised as the training exercises were performed in standing positions. A previous study by Whitney [58, 59] has suggested that DGI tends to improve after vestibular rehabilitation in patients with vestibular dysfunction. The results of our study are consistent with that claim. In addition, our data also reveal improved SOT test results as a consequence of the improved balance observed in the study patients.
The computer-aided exercise training system developed in this study was assembled by combining a PC with a simple Bluetooth transmission device. When connected to a Wiimote, this interactive training system can be used immediately after minimal adjustments. The cost of this system is relatively low, and it does not require the installation of any special connection cards. In contrast, commercially available assessment equipment is usually costly. Although other devices (i.e., linear potentiometers and optical gate-switching potentiometers) can be easily acquired, they cannot be easily installed and are not accurate in detecting head movements due to the structural limitations of the human head and neck. The limited angles and movement velocities supported by potentiometers do not meet the requirements for exercise. In addition, most commercially available assessment equipment requires circuit connections, which can result in unnecessary security concerns. In contrast, the light-weight training system developed in our study achieved rapid and practical wireless transmission. The important advantages of this interactive training system include convenience, mobility, safety and low cost.