A novel augmented reality approach enhanced visual acuity, contrast sensitivity, and ocular dominance in adults, including those with amblyopia, according to new research.

Investigators focused on the parvocellular (P) pathway, which is responsible for processing fine spatial detail and color, and is often impaired in amblyopia. They developed an altered reality system that selectively disrupted low spatial frequency (SF) visual input—processed by the magnocellular (M) pathway—by phase-scrambling it into flickering noise. High SF information was left intact to stimulate the P pathway. Investigators hypothesized that this method would reduce M pathway influence and allow enhanced P pathway function through disinhibition.

Participants were approximately 25 years old across the research, and those recruited for the third test were required to have at least 3 months of optical correction prior to the augmented reality (AR) training, among other inclusion criteria.

Test 1: Short-Term Visual Enhancement in Adults With Normal Vision

In the first test, 22 participants with normal vision underwent either 30 minutes or 4 hours of AR training. The 30-minute training resulted in a 4.4% increase in high SF contrast sensitivity, whereas the 4-hour training showed a 6.8% gain in high SF sensitivity, along with a 7.2% decline in low SF sensitivity. Improvement in high SF sensitivity correlated significantly with gains in visual acuity.

“These findings showed that short-term adaptation to the altered environment selectively enhanced high SF sensitivity in the parvocellular channel and improved visual acuity,” wrote lead author Yige Gao, of the Chinese Academy of Sciences in Beijing, and colleagues.

Test 2: Long-Term Dichoptic Training in Participants With Binocular Imbalances

Fifteen participants with interocular contrast imbalances underwent 5 days of 2-hour daily AR-based dichoptic training. In this protocol, the nondominant eye (NDE) received high SF images with low SF flickering noise and the dominant eye (DE) was shown high SF images at a low signal-to-noise ratio.

The NDE showed a significant increase in contrast sensitivity (P < .001), whereas the DE did not (P = .074). Binocular phase combination tests confirmed increased NDE dominance (P = .001). Improvements persisted 1.5 to 6 months post training, and steady-state visually evoked potentials confirmed enhanced NDE neural responses to high SF stimuli in early visual cortex.

“These findings provide direct neuroimaging evidence that NDE sensitivity to high SF naturalistic input was significantly enhanced in the early visual cortex,” noted Gao and colleagues.

Test 3: Home-Based AR Therapy in Adults With Amblyopia

The third test involved 26 adults with monocular amblyopia who used wearable AR glasses at home for 1 week (2 hours/day). The training protocol mirrored that used in test 2.

Best-corrected visual acuity improved significantly in the amblyopic eye under both crowded and uncrowded conditions (P < .001). NDE dominance improved (P < .001), and stereoacuity scores increased significantly (P < .001). Participants completed 102% of the prescribed training time with a 79% session completion rate.

“Compared [with] traditional training methods such as monocular patching or [visual perceptual learning],” the authors wrote, “our AR-based methods can have better treatment efficacy and patient compliance. In future studies, the image processing parameters, such as the cutoff spatial frequency, temporal frequency, and contrast ratio between different spatiotemporal frequency and ocular channels, can be individually optimized to further improve the training effect.”

Clinical Implications and Future Directions

This pathway-specific AR training approach holds promise for treating amblyopia beyond the critical period of visual development. The system may also be adaptable to other disorders involving magnocellular deficits, such as glaucoma and developmental dyslexia.

The researchers noted that the training capitalizes on both bottom-up (stimulus-driven) and top-down (cognitive) processes, and future optimizations could personalize parameters for greater efficacy.

They declared no conflicts of interest.

Source: Advanced Science