Feasibility of the Bioniko 3d-Printed Eye Model as a Training Tool for Internal Limiting Membrane Peeling

PURPOSE: To evaluate ophthalmology trainee's experience with BIONIKO 3D-printed eye model in internal limiting membrane peeling. METHODS: Ophthalmology trainees were randomly assigned to two groups, BIONIKO-first and Eyesi-first, performing internal limiting membrane peels on both models. Surgical performance was assessed via Eyesi score output, while an anonymous survey assessed model effectiveness. RESULTS: Thirty-nine trainees participated (20 BIONIKO-first, 19 Eyesi-first). The BIONIKO-first group had a higher mean Eyesi score (50.63 ± 31.68 vs. 37.53 ± 30.30, P = 0.497). The Eyesi-first group achieved a higher internal limiting membrane peeling rate inside the macula (95.05% ± 4.58 vs. 92.66% ± 10.75, P = 0.039). Completion time and forceps use were comparable. The BIONIKO-first group had fewer total spotted hemorrhages (2.4 ± 2.1 vs. 4.4 ± 3.6, P = 0.035), while macular and retinal injuries were similar.Subgroup analysis by training level showed that in the junior group, BIONIKO-first resulted in fewer hemorrhages ( P = 0.012), while Eyesi-first had higher internal limiting membrane peeling rate inside the macula ( P = 0.002). Among seniors, BIONIKO-first demonstrated a statistically lower area of macular injury (0.02 mm 2 ± 0.05 vs. 0.07 mm 2 ± 0.12, P = 0.020).Survey results favored BIONIKO for tactile feedback ( P = 0.002) and surgical realism ( P < 0.001), while Eyesi was preferred for procedural feedback ( P < 0.001). CONCLUSION: BIONIKO and Eyesi models offer complementary advantages. A hybrid approach may optimize skill acquisition and patient safety in vitreoretinal surgery training.