Augmented reality (AR) rehabilitation for homonymous hemianopia: ‘prism-free’ field expansion via smart glasses

To the Editor, Homonymous hemianopia (HH) is an acquired visual field impairment characterized by binocular loss of the same half of the visual field. HH results when the neural pathways carrying visual information from the eyes to the brain’s visual cortex, especially the optic radiations and occipital cortex, are injured so that the brain cannot process signals from the affected side of the visual field. This damage to the visual pathways is most commonly seen after ischemic cerebrovascular events[1]. Visual field impairment affects approximately 20–57% of stroke survivors. Diagnosis relies on automated perimetry and neuroimaging. Treatment primarily focuses on rehabilitation, compensatory approaches such as visual scanning exercises, and traditional prism lenses designed to improve the visual field and support daily activities[2]. Based on prior rehabilitative techniques, novel “prism-free augmented reality (AR) smart glasses” now offer real-time visual field expansion for patients with HH. These prism-free smart glasses work by capturing visual information from the impaired visual field and translucently superimposing it into the functional visual field without any delay. This technique operates as an adaptive visual field expansion tool, functioning like a real-time “picture-in-picture” display that augments the intact visual field without interfering with natural sight. In early trials of mixed reality navigation systems, patients with HH reported improved mobility using these augmentations. Despite being in initial trials, such interventions indicate potential for HH as well as other ocular deficits and low vision disorders[3,4]. Studies on prism-assisted rehabilitation for HH revealed mild-to-moderate functional benefit but limited cortical remapping. During early assessment, peripheral prism lenses expanded the binocular field by approximately 20–30°, associated with moderate functional mobility benefits, with neural adaptation being rare, and overall visual field recovery remained restricted[5]. In contrast, smart glasses using AR provide an adaptive, live visual overlay. This overlay is capable of stimulating cortical remapping by sending signals from the impaired field to the brain, activating the leftover visual pathways more effectively than conventional prism devices. Preliminary research shows that wearable AR systems expand the visual field and may improve navigation and perceptual awareness with minimal risk[6]. Despite promising initial outcomes, several challenges and limitations remain in the use of prism-free AR smart glasses for HH. First, the safety profile and long-term effects of constant use of AR smart glasses are not well established, with limited large-scale clinical trials evaluating adverse outcomes or cognitive load. Second, there is a lack of awareness and accessibility among clinicians and patients, which limits its use in regular rehabilitation programs. Additionally, most existing research uses small, heterogeneous groups with poor methodological rigor, making it difficult to generalize findings or compare devices accurately. Finally, the high cost and sophisticated technology of AR smart glasses and related software present significant challenges, particularly in low-resource settings, which reduces their widespread clinical use despite their potential advantages[4,7]. In conclusion, prism-free AR smart glasses show promising potential for improving functional vision in patients with HH. Further clinical trials involving larger and more diverse patient cohorts are crucial to enhance therapeutic outcomes. Continuous technological advancements, along with user-centered design improvements, are likely to boost performance. To fully realize their clinical benefits, future efforts should also focus on increasing awareness among healthcare providers and patients and providing subsidies to ensure widespread clinical adoption. This letter to the editor adheres to the Transparency in the Reporting of Artificial Intelligence in Research (TITAN) guideline[8].