Why a More Realistic Simulation Matters

Stargardt disease is a macular condition that can cause central vision loss—the loss of detailed vision used for tasks such as recognizing faces, reading, and seeing fine detail directly ahead. Yet central vision loss is often represented in virtual or extended reality (XR) with one simple image: a dark or blurry spot in the center of sight.

For many people living with Stargardt disease, that picture may not reflect what they actually experience. Clinical reports described in a new 2026 study suggest that central vision loss can involve far more than a fixed “black patch.” People may experience distortion of space, areas that seem perceptually absent, dimming, moving noise, temporal changes, or the brain’s apparent filling-in of missing visual information. These effects may vary substantially from one person to another, even among people with the same diagnosis.

The study introduces XR-VIEU, a gaze-contingent XR framework designed to model this diversity. “Gaze-contingent” means that the simulation responds in real time to where the user is looking, so the altered area of vision remains linked to the point of gaze rather than staying fixed on a screen.

Seven Building Blocks for Central Vision Loss

Rather than applying a single overlay, the researchers created seven independent visual-effect modules. These modules can be combined and adjusted to build different central vision loss profiles. The abstract identifies effects including spatial distortion, perceptual invisibility, dynamic noise, dimming, filling-in, and temporal disturbances; the system was designed to represent these clinically grounded phenomena in a flexible way.

This is an important design shift. A dark patch can be useful for communicating that central vision is impaired, but it can also imply that everyone sees the same sharply defined blind area. The new approach recognizes that visual loss may be experienced as altered, unstable, displaced, dimmed, or noisy vision—not simply as an empty region.

The framework was built in Unity’s High Definition Render Pipeline and includes formal mathematical descriptions of its effects. It offers more than 120 adjustable parameters for each eye, allowing researchers to alter the strength, shape, timing, and combination of simulated effects. That degree of control is intended to support individualized profiles rather than a one-size-fits-all depiction.

Two Stargardt Profiles, Two Very Different Experiences

To explore whether the system could reflect real-world experiences, the team conducted a profile reconstruction design study with two people who had late-stage Stargardt disease. The goal was not to establish a universal Stargardt vision pattern. Instead, the participants worked toward configurations that represented their own reported experiences.

The results were striking: although both participants had the same diagnosis and stage described in the paper, they needed substantially different combinations of modules. One participant’s profile was driven primarily by content displacement and dimming. In practical terms, this suggests that what they saw may have appeared shifted or altered in location, along with reduced brightness. The other participant’s profile relied mainly on dense fractal noise, a complex, texture-like visual disturbance.

These two cases provide early evidence that diagnosis alone cannot fully describe a person’s perceptual experience of central vision loss. Stargardt disease can share underlying clinical features across individuals while still producing highly individual visual experiences.

Fast Enough for Immersive Use

For an XR simulation to be useful, it must operate smoothly. Lag or low frame rates can be uncomfortable and can interfere with a user’s sense of immersion. The full seven-module XR-VIEU pipeline added 0.20 milliseconds of rendering time, or 2.9% overhead, while maintaining approximately 90 frames per second on a Varjo XR-4 headset at native resolution.

This performance result indicates that a detailed, composable simulation can be delivered in real time, at least on the hardware tested. That opens possibilities for research environments, professional education, and accessibility-focused design processes that need to account for gaze and motion.

Implications for Treatment Research and Care

XR-VIEU is not a treatment and does not test a therapy for Stargardt disease. Its potential contribution is different: it may offer a more precise way to understand, communicate, and study the functional impact of central vision loss.

For treatment research, personalized visual simulations could eventually help researchers discuss changes in vision with participants in more meaningful terms than a single generic image of a scotoma, or blind spot. In clinical education, they may help families, clinicians, designers, and researchers appreciate why two people with Stargardt disease can describe their vision differently. For accessibility-oriented design, developers may be able to test whether digital environments remain usable under a broader range of central vision experiences.

Looking Ahead

This publication is an early but meaningful step toward representing Stargardt-related central vision loss with greater fidelity. The evidence currently comes from two late-stage Stargardt participants, so larger studies will be needed to examine how well these profiles reflect the wider community and how they change over time.

Still, the central message is clear: central vision loss should not be reduced to a uniform black patch. By making visual effects modular, adjustable, and responsive to gaze, XR-VIEU creates a foundation for research and education that better reflects the individuality of lived experience with Stargardt disease.