International Guidelines for Photosensitive Epilepsy: Gap Analysis and Recommendations
J. Bern Jordan, Gregg C. Vanderheiden · 2024 · ACM Transactions on Accessible Computing · doi:10.1145/3694790
Summary
This paper provides the first comprehensive published comparison of five major international guidelines for limiting photosensitive epilepsy (PSE) risk from flashing content: UK Ofcom broadcast guidance, Japan's NHK/JBA guidelines, ISO 9241-391, ITU-R BT.1702, and W3C WCAG 2.x Success Criteria 2.3.1/2.3.2. These guidelines emerged after high-profile incidents—notably a 1997 Japanese Pokémon episode that caused 685 children to seek medical attention due to a 12 Hz red-blue alternation sequence. Photosensitive epilepsy is a reflex epilepsy where seizures can be triggered by specific visual stimuli, affecting approximately 1 in 4,000 people. The condition is diagnosed through intermittent photic stimulation (IPS) during EEG monitoring. Risk factors include flash magnitude (luminance difference), flash rate, area of flashing, and involvement of saturated red colors. All five guidelines address these factors but with varying thresholds and measurement approaches. The authors examine how technology has evolved since the mid-1990s when most thresholds were established. CRT televisions with inherent 50/60 Hz flicker have been replaced by LCD and OLED displays. Modern displays can be significantly brighter (smartphones reaching 1,000+ cd/m²), support high dynamic range (HDR) content, and cover wider color gamuts. VR/AR headsets present new challenges by filling nearly the entire visual field. The paper also notes the emergence of malicious "attacks" using flashing content against people with epilepsy on social media platforms.
Key findings
The authors identify several critical gaps and inconsistencies across guidelines: **Harmonization gaps**: While all guidelines limit flashing to 3 per second (Japan allows 5/second for moderate flashing), thresholds differ for luminance, area, and red flash definitions. The Japan guidelines are significantly more complex, with different rules for SDR/HDR content and scene changes. **Technology-driven changes**: - *Reduced risk*: Modern displays eliminate CRT's inherent flicker; higher refresh rates (120+ Hz) reduce visible flicker - *Increased risk*: Brighter displays (500-1,800 cd/m² vs. 200 cd/m² assumption), wider viewing angles with larger screens and VR, and high frame rate content (up to 120 fps) requiring new analysis approaches **New factors requiring guidance**: - *Flash transition duration*: No guidelines specify the timeframe for a transition. Authors propose 66 ms as a working value - *Synchronicity*: At high frame rates, transitions may span multiple frames across different screen areas. Authors propose 20 ms synchronicity window - *Saturated red definition*: ISO and WCAG 2.2 harmonized on a color-space threshold (R/(R+G+B) ≥ 0.8), but earlier WCAG versions used different calculations **Area thresholds**: WCAG uses a more conservative 0.006 steradian threshold (about 25% of the central 10° visual field), while other guidelines use 25% of total screen area. At typical viewing distances, WCAG leaves only ~10% of sensitive people at risk versus ~57% for 25%-of-screen thresholds. **VR/AR risk**: VR headsets fill 90°+ of visual field, meaning 25% of screen far exceeds the 0.006 steradian safe threshold. Authors recommend using WCAG's area threshold for VR content.
Relevance
This paper is essential reading for anyone developing content analysis tools, creating flashing content policies, or implementing WCAG Success Criteria 2.3.1/2.3.2. The detailed comparison tables and proposed definitions provide actionable technical guidance. **For content creators and platforms**: The paper explains why automated analysis tools may produce different results and helps understand which guideline to apply in different contexts (broadcast, web, games). The documentation of Twitter attacks targeting epilepsy organizations underscores the importance of content moderation for flashing media. **For tool developers**: The authors propose specific "working values" for parameters not currently defined in guidelines—66 ms flash transition duration, 20 ms synchronicity window, and methods for handling high frame rate content. These can inform development of next-generation analysis tools. The paper also surveys existing tools (HardingFPA, PEAT, PhotosensitivityPal, Flikcer) and their limitations. **For standards bodies**: The paper provides a framework for harmonizing guidelines across domains and suggests specific language for updated provisions. The authors have made their proposed definitions available on GitHub for potential incorporation into future standards. **For accessibility practitioners**: Understanding the rationale behind WCAG's PSE thresholds helps explain why content may fail analysis and how to evaluate risk when guidelines conflict. The discussion of dark-adapted viewing (cinema) and HDR content highlights edge cases where standard thresholds may be insufficient.
Tags: photosensitive epilepsy · seizures · flashing content · video safety · standards comparison · WCAG
Standards referenced: WCAG 2.0 · WCAG 2.1 · WCAG 2.2 · ISO 9241-391 · ITU-R BT.1702 · EN 301 549