iPad Eye Tracking Reproduces Clinical Grade Oculomotor Differences in Parkinson's Disease

iPad-based eye tracking could support Parkinson's disease (PD) screening and longitudinal monitoring by enabling objective, low-cost, portable assessment of oculomotor function. We previously validated an iPad-based eye-tracking system against the EyeLink 1000 Plus for temporal and spatial saccade metrics. Here, in a convenience sample of 19 healthy controls (HC) and 12 patients with PD, we recorded eye movements simultaneously with both devices during pro-saccade, anti-saccade (AS), memory-guided saccade (MGS), and self-generated saccade tasks. Across all pre-specified metrics, statistically significant PD-HC differences and null results were concordant between devices. In addition, saccade-level mixed-effects models showed small group × device interaction effects that remained below literature-based benchmarks for clinically meaningful PD-HC differences, indicating that iPad-based measurements preserved benchmark clinical-grade group-level effects. A compact three-metric iPad-based classifier comprising AS directional error rate, AS gain, and MGS gain supported strong subject-level PD-HC discrimination, with an area under the receiver operating characteristic curve of 0.98, sensitivity of 0.91, specificity of 1.00, and accuracy of 0.96. These findings support scalable tablet-based oculomotor assessment for PD-related screening and longitudinal monitoring.