Evaluation of a Novel Autostereoscopic Three-Dimensional System for Binocular Visual Function: A Comparative Validation Study

Background/Objectives: Autostereoscopic three-dimensional (3D) technology represents a promising tool for ophthalmic examinations. However, prior studies have largely focused on stereopsis alone in cooperative adults, leaving comprehensive validation across multiple binocular domains and diverse patient groups unexplored. This study introduces the head-to-head evaluation of a novel integrated, glasses-free autostereoscopic 3D system. The system could simultaneously assess four essential binocular functions (simultaneous perception, fusion, stereoacuity, and stereopsis), with a single automated workflow featuring real-time eye tracking. This study aimed to assess the comparative performance of the novel autostereoscopic 3D display in evaluating binocular visual function, as a preliminary potential screening tool benchmarked against conventional screening tests. Methods: A total of 232 participants (age: 3–60 years; male/female: 125/107) were recruited. All participants sequentially underwent binocular visual function assessments using autostereoscopic 3D technology, anaglyphic 3D technology (red–green glasses), and the Titmus stereopsis test. Pearson correlation analysis, Phi coefficient, Cohen’s Kappa and Bland–Altman plots were conducted to assess the consistency. Receiver operating characteristic curve analysis was employed to evaluate the screening comparative performance of autostereoscopic 3D technology, using the anaglyphic 3D technology and the Titmus stereopsis test as comparative reference standards. Results: Relative to anaglyphic 3D technology, agreement for simultaneous perception (φ = 0.527, κ = 0.425) and fusion (φ = 0.520, κ = 0.503) was moderate (all p < 0.001). Compared to the Titmus stereopsis test, agreement for stereoacuity levels (r = 0.779, κ = 0.659) and for stereopsis detection (φ = 0.624, κ = 0.619) was substantial (all p < 0.001). The areas under the curve for identifying abnormal binocular visual function using autostereoscopic 3D technology were 0.769 for overall binocular vision function, 0.747 for simultaneous perception, 0.730 for fusion, 0.846 for stereoacuity, and 0.810 for stereopsis (all p < 0.001). Sensitivities and specificities ranged from 52.8 to 81.7% and 77.9–93.1%, respectively. Conclusions: Autostereoscopic 3D technology demonstrated significant correlation and moderate agreement with traditional methods in assessing binocular visual function, particularly for stereopsis assessment.