An AI-Driven Micro-Randomized Intervention Framework for Reducing Digital Eye Strain and Dizziness during Prolonged Screen Use

Extended exposure to digital screens has been associated with an increasing incidence of digital eye strain (DES), dizziness, and visual discomfort. Existing mitigation strategies, such as scheduled breaks and blue-light filtering, are largely static and do not adequately account for individual variations in fatigue response. In this work, an AI-based adaptive system is presented that monitors blink rate patterns, gaze stability, posture-related variations, and screen exposure duration to identify early signs of visual fatigue. When fatigue is detected, a micro-randomized and gamified eye-exercise module provides short, personalized interventions designed to alleviate strain and improve visual comfort. The proposed approach was evaluated through a controlled study involving 100 participants and compared with conventional break-based methods using standardized eye strain and dizziness questionnaires, together with AI-derived fatigue indicators. The findings indicate a measurable reduction in digital eye strain and dizziness, along with improved user engagement and comfort. Variability across participants was observed, suggesting that adaptive and personalized intervention strategies are particularly important for managing prolonged screen use. Overall, the study supports the potential of AI-driven micro-randomized interventions for enhancing visual wellness in screen-intensive environments.