A Novel Eye Tracking-Based Gamified Assessment of Contrast Sensitivity Function in Children: Prospective Development and Reliability Study.

Background: Reliable assessment of visual function in young children remains a challenge. Contrast sensitivity function (CSF) is a sensitive and fundamental index of visual performance, yet existing pediatric CSF assessments lack objectivity and adaptability. To bridge this methodological gap, we developed a novel eye tracking-based gamified contrast sensitivity function (ETGCSF) tool that integrates gaze-based detection with interactive gameplay to objectively quantify CSF in an engaging and child-centered manner. Objective: This study aimed to (1) establish the feasibility and test-retest reliability of the ETGCSF tool in preschool-aged children and (2) evaluate whether optimization using adaptive algorithms and enhanced gamification elements could improve test efficiency while maintaining reliability. Methods: This was a prospective study with 2 sequential cohorts. A total of 80 Chinese children aged 3 to 6 years were pragmatically recruited from Zhongshan Ophthalmic Center between May 2021 and July 2023. On the basis of timing of data collection, 35% (28/80) of the children were included in experiment 1 (mean age 5.24, SD 0.15 years), and 65% (52/80) were included in experiment 2 (mean age 4.76, SD 0.11 years). Children completed 2 runs of ETGCSF test. Experiment 1 used the baseline ETGCSF protocol, and experiment 2 used the optimized protocol. Primary outcomes were test-retest reliability of the area under the log contrast sensitivity function curve (AULCSF) and CSF acuity, reported as intraclass correlation coefficients (ICCs) with 95% CIs. Results: In experiment 1, the ETGCSF tool showed strong reliability, with ICCs of 0.890 (95% CI 0.741-0.951) for AULCSF and 0.890 (95% CI 0.763-0.949) for CSF acuity. The median test duration was 482 (IQR 451-569) seconds. In experiment 2, the optimized ETGCSF reduced median test duration to 241 (IQR 189-296) seconds (P<.001) while maintaining comparable reliability. AULCSF estimates varied by 0.03 log units across 2 runs (95% CI -0.51 to 0.57; t51=0.749; P=.46), with an ICC of 0.851 (95% CI 0.740-0.914; P<.001) that was not significantly different from that of experiment 1 (z=0.660; P=.51). Similarly, CSF acuity estimates varied by 0.004 log units (95% CI -0.33 to 0.32; t51=0.192; P=.85), with an ICC of 0.832 (95% CI 0.708-0.904; P<.001), also comparable to that of experiment 1 (z=-0.925; P=.36). Conclusions: This study introduces a paradigm shift in pediatric visual assessment by leveraging objective eye tracking and gamified engagement to transform contrast sensitivity testing into a scalable, child-friendly process. The ETGCSF tool demonstrated strong reliability and markedly improved efficiency in assessing CSF in preschool children aged 3 to 6 years. These findings support ETGCSF as a promising tool for real-world clinical practice, and its modular design holds potential for future adaptations ranging from streamlined rapid screening in very young children to full CSF profiling for research.