Aim To evaluate the performance of a parallel robotic system by comparison with the conventional manual procedure using an animal model.

Methods A new parallel robotic system was developed that features a small cylindrical manipulator (base diameter 76 mm, height 240 mm). The performance of the new system was evaluated for its capability to assist in retinal vessel microcannulation. The test scenario was as follows: (1) introduce the microcannula into a harvested porcine eye attached loosely on the orbital fossa of an artificial face model through a 20G scleral port at the pars plana; (2) cannulate the retinal vessels (inner diameter 60–80 μm); and (3) inject indocyanine green dye into the eye endovascularly. The success rate and procedure quality of the robotic system were evaluated by comparison with the conventional manual procedure.

Results Retinal vessel microcannulation and dye injection were achieved by the robotic system twice in four attempts, and by the conventional manual procedure either not at all or incompletely in all six attempts. Dye leakage was not observed with the robotic system, indicating that microcannulation was minimally invasive; in contrast, dye leakage was always observed with the manual procedure.

Conclusions The new system is more accurate than the conventional manual procedure for the tests on a porcine eye model.