Currently, rehabilitation methods for patients with tunnel vision are still relatively limited. Existing VF expansion aids include reflectors, prisms, inverted Galilean telescopes, etc. However, the expansion of VF of these aids is minimal, and there are shortcomings, such as dispersion, imaging deviation, and poor patient adaptation [4,5,6]. HMDs avoid these problems. They can significantly expand VF and adjust the display mode according to the needs of different patients, which is expected to provide adequate assistance for patients with severe tunnel vision [11,12,13]. There are several augmented reality devices available on the market, including eSight. Studies have shown that eSight can improve visual function and increase the quality of life of patients with central vision loss [14, 15]. However, there are few studies of HMD applied to patients with PVL (especially tunnel vision), and there is a lack of research about the effect applied in the activities in real environments. In addition, the previously reported VF expansion HMD visual aids all used the contour display mode. The Acesight AR studied in this paper is a new head-mounted visual aid that uses a reduced image display mode, which can provide more information than a contour. To the best of our knowledge, no studies have reported the clinical application effect of this type of visual aids mode.

More than half (54%) of the 57 patients in this study found Acesight helpful. Young-to middle-aged patients were likelier to learn and understand the use patterns of the visual aids and thought it useful. Other studies have found that young patients were more receptive to VR HMD visual aids [16, 17], and this study demonstrates that younger patients were more receptive to AR HMD visual aids as well.

In this study, Acesight significantly enlarged the horizontal and vertical VF of patients with tunnel vision. In previous research, HMD visual aids can expand the VF to a radius of 10° to 15° [18, 19]. The horizontal VF diameter in this study was expanded to about 35°, and the vertical VF diameter was extended to about 30°, which was better than in the previous study. In this study, the expansion of the horizontal VF diameter was greater than the vertical VF. This may be because the VF angle of Acesight image acquisition is 25° radius in the horizontal direction and 15° in the vertical direction, and the horizontal direction is larger than the vertical direction. In addition, our study showed that Acesight was more effective at expanding VF in patients with a VF diameter of < 20° than in patients with a VF diameter of ≥ 20°. This indicates that Acesight has a better effect on VF enlargement for patients with more severe VF damage.

Patients with tunnel vision are prone to danger due to narrowed VF and difficulty noticing pedestrians or vehicles passing on the side of the walking path. In addition, the patient’s mobility can be reduced because of a reduced ability to spot obstacles, resulting in navigation difficulties. In this study, after wearing Acesight, patients could detect people moving sideways earlier and avoid the risk of collision to some extent.

The HMD vision aids reported by Peli et al.[20] for patients with PVL compressed the contour information of the larger field of view and superimposed it in the patient’s central field of view. Compared with no enhancement, minification with contour images also improved visual search performance when the participants’ original VF was not too limited, and auditory cues decreased search time for all the participants by 54% on average [21]. However, the visual field search time was prolonged in patients with severe tunnel vision (field diameter < 10°) [22]. Acesight uses AR display technology in this study to compress information from a wide visual field into the patient’s residual visual field. There was no significant difference in the search ability of patients with tunnel vision with and without Acesight. These different results may be due to the different display modes and testing methods of visual aids. The technique used by Peli et al. was to find a particular low-contrast letter on a white screen, while our study was in a real environment. In addition, due to the small number of participants included (12 in the former study and 57 in this study), the study’s results were susceptible to the influence of the participants. Patients with tunnel vision mainly search for objects by rolling their eyes and turning their heads. In addition to expanding the VF, adding contours to objects and improving contrast could theoretically help in searching. Which display mode is more helpful in improving the search ability of patients with tunnel vision needs further study.

Our study found that Acesight may reduce the speed of independent walking in patients with tunnel vision in the first use. Similar results have been found in previous studies of other VF enlargement aids [12, 23]. When wearing an HMD, patients were more hesitant to walk independently, thus decreasing their walking speed. VF expansion strategies include compressing visual information, image superposition, etc. Still, these strategies will lead to changes in the patient’s visual perception mode, and it is often difficult for patients to adapt quickly to the new visual perception mode in first use. Acesight compressed visual information into the patient’s residual VF in this study. At the same time, the reduction of the image would change the spatial perception, and the distance perception of patients would worsen. In addition, patients with tunnel vision often compensate for visual field loss by rolling their eyes. When patients wear HMDs, they need to turn their heads to change their field of vision, which is different from how they are accustomed to it. The reduction in mobility ability may relate to the limitations imposed on the field that can be scanned by eye movements alone (without head movement) when wearing the device. Due to the minification factor, training may be necessary to gain a veridical perception of visual direction and correspondence between the real world and the displayed image. Other studies have shown that a period of rehabilitation guidance and training may help improve the acceptance of HMD visual aids [24, 25]. Further study is needed to determine whether the effect will be enhanced after a more extended period of training and use of the device in realistic outdoor settings.

Our study found that VA was decreased with the Acesight device in VF expansion mode compared to when not worn, due to the reduction of the image. When patients with tunnel vision wear Acesight during movement, the VF expansion mode can help detect people and vehicles on the side earlier, reducing the risk of collision. According to our study, more patients with VA (logMAR) ≤ 0.5 thought the Acesight helpful (Fig. 3e). We thought that the VF expansion mode is more suitable for patients with better VA, which is necessary to process the reduced image display. When the patient uses it in static scenarios, they can switch to magnification mode, and the Acesight will display an enlarged image so that the patient can see the details clearly and avoid the inconvenience caused by vision loss.

Subjects were allowed to comment freely on Acesight and how they thought it could be improved. Despite the open-ended nature of the question, several common themes emerged. The improvement directions of HMD VF expansion aids in the future include: (1) Improve clarity to expand VF while minimizing the loss of clarity; (2) Improve portability, such as adopting a glasses-like appearance, which is convenient for patients to use outdoors; (3) Expand the camera range and further improve the ability of VF expansion; (4) Increase the anti-shake function to reduce image jitter caused by autofocus during head movement and avoid dizziness, fatigue, and other problems caused by it.

There were several limitations to this study. The tests in our study were all performed without the device first and then with the device. Several hours of training between tests may have diminished the memory effect, but it could not be avoided altogether. In this study, patients only adapted to Acesight for several hours, and some patients’ walking ability was improved after training. Further study is needed to determine whether the effect will be enhanced after a more extended period of training and use of the device in realistic outdoor settings.