Scientists have developed an autonomous robotic system that could revolutionise delicate eye surgery, offering new hope for patients with retinal vein occlusion (RVO).
RVO is a serious condition in which a retinal vein becomes blocked, potentially leading to vision loss. Retinal vein cannulation (RVC) requires surgeons to insert a tiny needle into the blocked vein with high precision, delivering clot-dissolving drugs or medications that control abnormal blood vessel growth.
Given that retinal veins targeted for cannulation are about as thick as a human hair, manually inserting a needle with such precision can be challenging.
Researchers at Johns Hopkins University have now developed an autonomous robotic system capable of reliably performing these procedures, ensuring that needles are inserted correctly without damaging the patient’s retina.
This robotic system, described in a paper in Science Robotics, is guided by deep learning algorithms that analyse images from a surgical microscope and cross-sectional scans of eye tissue obtained using optical coherence tomography (OCT).
Developed by Peiyao Zhang and his colleagues, it integrates computational techniques for robot control with high-resolution imaging and deep learning algorithms.
The system consists of two robots – steady-hand-eye robots – designed to perform retinal microsurgeries. Each robot holds a tiny needle and a surgical tool.
This hardware is combined with three deep learning algorithms trained to track the needle’s movement and plan the robot’s actions to ensure the needle is correctly inserted into the affected retinal veins.
The researchers have so far tested their system on pig eyes that were either still or moved periodically, mimicking the breathing motions of live pigs or humans.
In the team’s experiments, the robotic system successfully completed RVC on 90% of still pig eyes and 83% of moving pig eyes.
Notably, the system could also reliably detect when the needle made contact with and entered a retinal vein.
The paper suggests that robotic systems could perform RVC procedures, but further testing in animals and humans is needed before clinical use.
Zhang’s robotic system might benefit eye surgeons and expand RVO treatment options. The team plans further studies to validate their system.
